Tai Chi cognitive intervention method and system based on exercise psychology

Abstract

The invention belongs to the technical field of exercise psychology and cognitive behavior intervention, and particularly discloses a shadowboxing cognitive intervention method and system based on exercise psychology. The method comprises the steps that historical cognition and emotion data of an intervention object are collected, multi-dimensional state evaluation and grouping are carried out in combination with action execution performance, and then a personalized shadowboxing training scheme and a psychological intervention strategy are formulated. The system dynamically adjusts an intervention path through multi-source data acquisition, action recognition, emotion perception and an intervention effect feedback mechanism, and realizes continuous optimization of a cognitive state. The method and the system can be widely applied to psychological rehabilitation, educational intervention and emotion regulation scenes, and have the characteristics of high intelligent degree, high intervention precision and good adaptability.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of sports psychology and cognitive behavior intervention, and specifically discloses a Taijiquan cognitive intervention method and system based on sports psychology. BACKGROUND

[0002] As a traditional health-preserving exercise form, Taijiquan combines the coordination and unity of breathing, movement and consciousness, and has unique advantages in enhancing physical fitness and improving psychological state. In recent years, as mental health problems have been increasingly valued, Taijiquan has gradually been included as one of the auxiliary intervention tools in the fields of cognitive intervention, emotional regulation and psychological rehabilitation. At the same time, the development of sports psychology has also given traditional exercise methods a more scientific and individualized guidance path, prompting the integration of traditional Taijiquan and modern psychological intervention technology to become a new research hotspot.

[0003] Existing intervention methods are mostly based on traditional Taijiquan routine training, supplemented by general cognitive education, behavior guidance or group psychological counseling, trying to improve individual symptoms such as anxiety, depression, attention deficit or cognitive decline. However, most of these methods lack a deep combination of individual cognitive structure, psychological regulation mechanism and movement learning rules, the intervention process lacks a dynamic evaluation mechanism, the feedback means is not timely, and few of them integrate visual evaluation tools, movement behavior modeling and individual psychological characteristic analysis modules. In addition, existing systems usually focus on the execution of standard movements of physical exercise, and pay little attention to the whole process interaction of movement imagery, emotional arousal and cognitive participation, resulting in unstable intervention effectiveness and difficulty in adapting to the individual needs of different groups.

[0004] Therefore, it is urgent to build a new intervention mode that integrates the principles of sports psychology, emphasizes cognitive regulation and body movement coordination, in order to improve the application depth and adaptability of Taijiquan intervention in the field of mental health in a more systematic and intelligent way. SUMMARY

[0005] In view of the above problems, the purpose of the present application is to provide a Taijiquan cognitive intervention method based on sports psychology, comprising the following steps:

[0006] S1, collecting historical data of the intervention object, extracting cognitive state indicators, and preliminarily grouping the intervention object;

[0007] S2, standardizing the cognitive state indicators of the intervention object, calculating the cognitive state score by weighting method according to the set weight factor, and adjusting the grouping according to the score result; wherein the weighted items of the cognitive state score include the psychological test score, the cognitive state score and the behavior stability score;

[0008] S3. Obtain Tai Chi learning videos of the intervention subjects, and conduct content analysis in conjunction with the corresponding cognitive state scores to determine whether the movement execution status is within the standard range.

[0009] S4. Obtain the emotional index of the intervention subjects, and integrate and analyze the emotional index and cognitive state score according to the preset scoring weights and matching rules to complete the final grouping.

[0010] S5. Develop group intervention plans, implement Tai Chi training and psychological intervention, collect and analyze data during the implementation process in real time, evaluate the completion of the intervention, and dynamically adjust the intervention plan until the preset intervention goals are achieved.

[0011] As a preferred technical solution, the determination of whether the action execution state is within the standard range is completed by extracting the skeleton key point coordinate sequence of the action key frame and performing action matching and similarity analysis with the standard action template.

[0012] As a preferred technical solution, in step S1, the cognitive state indicators include focus, comprehension, and memory. Collecting historical data from the intervention subjects includes the following steps:

[0013] Collect the intervention subjects' historical psychological test scores, learning duration, basic assessment scores, number of motor errors, and emotional state fluctuation values;

[0014] The historical psychological test scores, basic assessment scores, and learning time were standardized to obtain three sets of standardized basic vectors.

[0015] The above basic vectors are set as a group, and the number of movement errors and the value of emotional state fluctuations are set as compensation vectors. The cognitive state index is obtained by solving the weighted distance between the basic vectors and the compensation vectors.

[0016] The formula for calculating the weighted distance is:

[0017]

[0018] in, , , The corresponding term of the basic vector, , , For the corresponding term of the compensation vector, , , The correlation weights between each dimension and the cognitive state score are given, and .

[0019] As a preferred technical solution, step S4, obtaining the sentiment index includes the following steps:

[0020] Facial expression images and physiological signal data of the intervention subjects were collected, and expression parameters were extracted from the expression images by grayscale conversion and key feature point detection algorithms.

[0021] The physiological signals were processed by bandpass filtering to obtain heart rate variability and skin conductance intensity.

[0022] The facial expression parameters are paired with the feature vectors of physiological signals using a time synchronization mechanism, and their weighted average value is calculated.

[0023] By setting a threshold, signals in a neutral state are identified and eliminated. The remaining signals are used to calculate the current sentiment index.

[0024] As a preferred technical solution, step S5, developing a group intervention plan includes the following steps:

[0025] Establish training module libraries and intervention strategy module libraries;

[0026] Based on the final grouping results, the corresponding training modules and intervention strategies are extracted from the above module library and combined to generate an intervention plan.

[0027] The training module includes a beginner module, an intermediate module, and an advanced module; the intervention strategy module includes a cognitive guidance module, a feedback reinforcement module, and an emotion regulation module.

[0028] As a preferred technical solution, during the implementation of the intervention plan, execution process data is acquired after each set of training modules is completed. The execution process data includes the execution cycle, completion rate, and feedback difference. The feedback difference is the difference between the final similarity of the executed action and the standard similarity of the reference action.

[0029] As a preferred technical solution, in the process of using execution process data to calculate the intervention completion rate, the intervention completion rate is set as the objective function, an iterative process based on Bayesian optimization is used for parameter updates, and dynamic constraints are set on the execution cycle; the expression of the objective function is:

[0030]

[0031]

[0032] in, To find the maximum expected corresponding combination that is desired in the Bayesian optimization iteration. To optimize the function The expected output value, For mathematical expectation operators, For the execution cycle, To achieve the average final similarity, This is the feedback difference; Let be the maximum solution function, representing the solution that makes the function maxima ≤ 0. The largest expected output value The combination; , and Optimization function The sub-item strength adjustment coefficient is set based on empirical values. This is the standard execution cycle.

[0033] This invention also provides a Tai Chi cognitive intervention system based on sports psychology, used to perform the method, including:

[0034] The assessment module is used to collect historical data of intervention subjects, extract cognitive status indicators, and perform preliminary grouping;

[0035] The judgment module is used to obtain cognitive state scores and determine whether the intervention subject's action execution is standardized.

[0036] The integration module is used to obtain the emotion index, establish a correlation model between emotion and cognitive state, and complete the final grouping;

[0037] The intervention module is used to develop intervention plans and implement Tai Chi training and psychological intervention, and output the intervention completion rate at each stage;

[0038] The adjustment module is used to dynamically adjust the intervention plan based on the progress of the intervention until the intervention goal is achieved.

[0039] As a preferred technical solution, the evaluation module is also used to collect historical psychological test scores, learning duration, basic assessment scores, number of motor errors, and emotional fluctuation values, perform standardization processing, and construct basic vectors and compensation vectors. Cognitive state indicators are then calculated using a weighted distance calculation, the formula for which is:

[0040]

[0041] in, , , The corresponding term of the basic vector, , , For the corresponding term of the compensation vector, , , The correlation weights between each dimension and the cognitive state score are given, and .

[0042] As a preferred technical solution, it also includes a feedback module, which is used to collect the facial expressions and physiological signals of the intervention subjects and input them into the emotion analysis algorithm to output the corresponding emotion index;

[0043] The judgment module integrates the emotion index and cognitive score to determine the consistency of the intervention. If there is a discrepancy, the groups are regrouped and a warning signal is issued.

[0044] As a preferred technical solution, when formulating an intervention plan, the intervention module calls the training module library and the intervention strategy module, selects the appropriate module according to the final grouping, and combines them to generate multiple sets of intervention plans for intervention execution and adjustment.

[0045] The beneficial effects of this invention are as follows:

[0046] This invention proposes a Tai Chi cognitive intervention method and system based on sports psychology. It can simultaneously consider the standardization of Tai Chi movement training and the psychological stability of the intervention subjects in sports cognitive intervention, and realize the construction and execution of individualized and dynamic intervention paths. This invention obtains the historical data of the intervention subjects, extracts cognitive state indicators, and preliminarily groups the intervention subjects according to these indicators. Then, it establishes an evaluation model, adjusts the grouping results according to the cognitive state scores, and on this basis, obtains the Tai Chi learning videos of the intervention subjects to judge the standardization of the performed movements.

[0047] After obtaining the emotional index of the intervention subjects and establishing a correlation model between emotion and cognition, the grouping results are optimized, a grouping intervention plan is formulated, and a combination of Tai Chi training and psychological intervention is implemented. Data on the execution process is collected and analyzed in real time. By calculating the intervention completion rate, the intervention plan is dynamically adjusted until the preset intervention goals are met. This invention ensures the scientific nature, pertinence, and stability of the intervention while fully considering the individual differences of the intervention subjects, and improves the overall effect of cognitive intervention. Attached Figure Description

[0048] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0049] To enhance understanding of the present invention, the present invention will be further described in detail below with reference to embodiments. These embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0050] Example 1: A Cognitive Intervention Method for Tai Chi Based on the Fusion of Sports Psychology Features:

[0051] This embodiment is based on the principles of sports psychology and integrates a multi-dimensional cognitive state assessment mechanism, emotion index acquisition technology, video action recognition method, and nonlinear optimized intervention scoring function to achieve personalized generation and dynamic adjustment of intervention plans.

[0052] This embodiment used 40 university students aged 18 to 25 as a pilot sample. The experimental period was set at 28 days, divided into an initial assessment period (3 days), an intervention implementation period (21 days), and a follow-up assessment period (4 days). The entire process followed... Figure 1 The steps shown are specified and should be performed as follows:

[0053] S1. Historical Data Collection and Preliminary Grouping:

[0054] Before the intervention began, historical cognitive data were collected from each participant, including:

[0055] Study duration (unit: minutes / day, average of the last 30 days);

[0056] Historical psychological test scores (standardized to a percentage).

[0057] Basic psychological assessment results (including attention index and short-term memory score);

[0058] The frequency of errors in movements during past participation in physical education courses;

[0059] Emotional fluctuation records (collecting morning and nighttime facial expression scores, heart rate changes, etc., and calculating the standard deviation).

[0060] The above data is normalized, and the feature vector is constructed as follows:

[0061] Fundamental vectors ,in:

[0062] Standardized psychological test scores;

[0063] Basic assessment overall score;

[0064] Average study time.

[0065] Compensation vector ,in:

[0066] : Frequency of errors per unit time;

[0067] Emotional fluctuation value;

[0068] Instantaneous heart rate fluctuation range.

[0069] Wherein, the base vector and the compensation vector are both three-dimensional vectors. The weighted Euclidean distance between the two vectors is calculated as the cognitive state score D. The numerical expression of the weighted Euclidean distance is as follows:

[0070] in, , , Participants were initially divided into three groups based on their scores, from highest to lowest: a high cognitive state group, a medium cognitive state group, and a low cognitive state group.

[0071] S2: Cognitive Status Score Calculation and Group Adjustment

[0072] Based on the initial grouping, a cognitive scoring system is further constructed. Combining the aforementioned standardized vectors, a linear weighting method is used to generate a scoring function, resulting in a cognitive state score, Score, which satisfies:

[0073]

[0074] in, , , In addition, the results of subjective interviews were combined with manual review of the marginal group members to make minor adjustments to the final grouping.

[0075] S3: Video Action Recognition and Execution Status Assessment

[0076] To capture the performance status of the subjects during Tai Chi practice, daily training sessions were recorded using a camera, and key skeletal point sequences were extracted using a deep pose recognition model. Similarity was calculated using key movements from the standard 24-form Tai Chi, such as "Starting Posture," "Single Whip," and "Wild Horse Parts Its Mane," as comparison standards. Specifically, this included keyframe extraction of the movements. The video data of the intervention subject is input into a human pose recognition model based on a deep neural network (such as HRNet, OpenPose or MoveNet), and the coordinate sequence of the skeleton key points of consecutive frames during the training process is extracted, including at least 18 key points such as the torso, limbs, and joint nodes.

[0077] Standard action template construction: The system has a pre-set database of standard movements for the 24-form Tai Chi Chuan, containing the temporal skeletal trajectory, key posture transition paths, and rhythm weight information for each movement. Each movement template is labeled with structural features such as the start and end times of the movement, rhythm change points, force direction, and body rotation amplitude.

[0078] Action matching and similarity analysis: After standardizing the collected skeleton data, the Dynamic Time Warping (DTW) algorithm and cosine similarity calculation are used to match the actual action sequence of the intervention object with the standard action template to obtain the action similarity score S, with a value range of [0,1]. The higher the value, the closer the action is to the standard trajectory.

[0079] Threshold adjustment based on cognitive state scores: If the cognitive state score is set to C (output from steps S1–S2), then the system adjusts the similarity judgment threshold T based on the C value: When C is high (e.g., C>0.8), T is set to a higher value (e.g., T=0.85) to require more standardized execution of the action; When C is moderate (e.g., 0.5≤C≤0.8), T is set to the middle value (e.g., T=0.75). When C is low (e.g., C < 0.5), T is set to a lower value (e.g., T = 0.65) to allow for a larger range of deviation. This flexible threshold mechanism based on cognitive state adaptation can avoid putting excessive pressure on individuals with high cognitive load, while maintaining intervention accuracy for those with high cognitive abilities.

[0080] Determine the standardization of the action: If the similarity score S of a certain action is greater than or equal to the adjusted threshold value T, then the action is determined to be in a normal state; otherwise, it is determined to be in a non-normal state, and its error peak frame and error cumulative value are marked for subsequent feedback module analysis.

[0081] A combination of Dynamic Time Warping (DTW) and cosine similarity is used to extract the final similarity of actions for each training session. The range is between 0 and 1. The training period is also recorded. (in minutes) and feedback difference (The mean square error of the deviation between the system's real-time feedback and the actual behavior).

[0082] S4: Sentiment Index Collection and Fusion Analysis

[0083] Within 5 minutes after daily training, the following data were extracted using a facial expression recognition algorithm and a finger pulse heart rate sensor: facial muscle tension; heart rate variability (HRV); respiratory rate; after time-aligned processing, an emotion index was synthesized using the following formula:

[0084]

[0085] Wherein, λ1=0.5, λ2=0.3, λ3=0.2; S1 is the facial expression score, which is the normalized value of the sample mean squared error of facial muscle tension; S2 is the normalized value of heart rate variability; and S3 is the respiratory stability score, which is the normalized value of the sample mean squared error of respiratory rate. A discrimination threshold of 0.45 was set, and those below this value were considered to be in an unstable emotional state. Cross-analysis was performed in conjunction with cognitive state scores to reconfirm the final intervention group.

[0086] S5: Develop and iteratively adjust intervention plans.

[0087] Based on the final grouping results, the system retrieves the corresponding Tai Chi training content and psychological intervention strategy modules from the intervention module library:

[0088] The beginner group combines "5-style Tai Chi + breathing regulation + cognitive guidance";

[0089] The intermediate group implemented "10-style Tai Chi + imagery training + emotional response diary";

[0090] The advanced group used "24-style Tai Chi + scenario meditation + self-feedback reconstruction".

[0091] Each training module records the following parameters: For the execution cycle, To achieve the average final similarity, The feedback difference is obtained through the following nonlinear function. Calculate the intervention score for this training session:

[0092]

[0093] When a training score falls below 0.4, or when the training trend declines for three consecutive training sessions, the system automatically adjusts the intervention module (reduces the difficulty or changes the strategy) and marks it as "abnormal status" to enter the auxiliary monitoring process.

[0094] For nonlinear functions Perform Bayesian iteration to find the combination corresponding to the maximum expectation. :

[0095]

[0096] in, To find the maximum expected corresponding combination that is desired in the Bayesian optimization iteration. To optimize the function The expected output value, For mathematical expectation operators; Let be the maximum solution function, representing the solution that makes the function maxima ≤ 0. The largest expected output value The combination; , and Optimization function The sub-item strength adjustment coefficient is set based on empirical values. This is the standard execution cycle;

[0097] Furthermore, after the intervention period ends, the following assessment methods are used to determine the intervention effect:

[0098] The similarity of movements in the complete Tai Chi routine has improved after comparison.

[0099] Comparison of cognitive assessment scores before and after;

[0100] Sentiment index mean and fluctuation range;

[0101] Subjective satisfaction survey.

[0102] Taking a participant as an example, the average similarity score before the intervention was 0.68, which increased to 0.89 after the intervention. The average intervention score rose from 0.45 to 0.82, the cognitive assessment score increased by 12 points, and the amplitude of emotional fluctuations decreased by 37%, showing a good intervention response.

[0103] This embodiment comprehensively verifies the feasibility and effectiveness of the method in real-world populations. Through multi-dimensional data fusion, dynamic strategy adjustment, and a non-linear scoring mechanism, individual differences can be accurately identified and effective interventions can be implemented. Furthermore, the construction of the scoring function considers behavioral performance, error compensation, and execution regularity, giving the intervention plan a high degree of adaptability and dynamic optimization capability, significantly outperforming traditional static grouping and quantitative evaluation systems.

[0104] Example 2: A Tai Chi Cognitive Intervention System Based on the Fusion of Sports Psychology Features:

[0105] This embodiment provides a Tai Chi cognitive intervention system based on sports psychology. It aims to systematically, progressively, and personally enhance the cognitive abilities of individuals with weaker cognitive states by integrating multi-source data collection, real-time motion recognition, emotional feedback perception, and intelligent intervention strategy recommendations. This system is widely applicable to university mental health intervention centers, rehabilitation training institutions, and home-based remote cognitive rehabilitation scenarios.

[0106] The system consists of five core functional modules: assessment module, judgment module, integration module, intervention module, and adjustment module. These modules work together through a data bus and scheduling control logic to achieve a closed-loop information system for the entire intervention process.

[0107] I. Evaluation Module:

[0108] The assessment module is responsible for collecting various basic data points from the intervention participants during the intervention initiation phase, and for conducting a preliminary assessment and grouping of their cognitive status. Data collection methods include:

[0109] 1) Collect historical academic performance, average study time, and basic psychological assessment results through the client-side questionnaire module;

[0110] 2) Collect the error frequency from the user's past exercise training records through the motion history record module;

[0111] 3) Synchronously obtain the emotional fluctuation indicators of the past period through the emotion recording interface, including daytime facial expression recognition and nighttime heart rate recording;

[0112] 4) After the collected data is standardized on the backend server, the feature vector assembly logic is called to format the structured data into two groups: cognitive basis vectors (such as grades, assessments, duration) and emotion / behavioral compensation vectors (such as error rate, volatility, etc.).

[0113] Based on this, the assessment module divides the subjects into three initial intervention groups: high-sensitivity response group (Group A), moderate adaptation group (Group B), and low response delay group (Group C), and outputs the results to subsequent modules.

[0114] II. Judgment Module:

[0115] The judgment module is responsible for acquiring information on the movement state of the intervention subject during Tai Chi training from the video recognition and motion capture terminals, and evaluating whether it conforms to the execution standards corresponding to the predetermined cognitive state. Specifically, this includes:

[0116] 1) The training process of the intervention object is collected in real time by camera, and the key frames of the action skeleton are extracted by the embedded human key point recognition algorithm.

[0117] 2) Compare the standard procedures of each movement in the standard Tai Chi movement database, and use the inter-frame movement similarity measurement model to calculate the execution deviation;

[0118] 3) Analyze the current action matching degree, training rhythm and system prompt response time to determine whether there are any abnormalities in the object's actions (such as delayed start, repeated errors, disordered rhythm, etc.).

[0119] 4) The judgment results are divided into three categories: standard action, adjustable action deviation, and abnormal action that needs adjustment.

[0120] This module is also responsible for real-time statistics on the completion status of the current training round and outputting it in a structured manner to the integration module and the adjustment module.

[0121] III. Integration Module:

[0122] The integration module is responsible for the task of fusing and judging emotional state and cognitive executive state, and is the core part of realizing multimodal data linkage analysis.

[0123] 1) This module accesses emotion recognition data, including facial expression recognition (using a multi-region convolutional feature fusion algorithm), voice emotion analysis (such as intonation changes during self-talk in training), and physiological signals (such as HRV and skin conductance).

[0124] 2) Align the above data to the current training time period through a time synchronization mechanism, and smooth out extreme fluctuations in short time windows;

[0125] 3) Based on the coupling between the execution of training actions and emotional responses, construct a "cognitive-emotional consistency matrix" to determine whether the current intervention subject is undergoing cognitive load training in an emotionally stable state;

[0126] 4) If the system determines that there is an “emotion-action inconsistency”, that is, during the training process, the action is passively executed but the expression is suppressed, the heart rate is increased, etc., then the state is marked as a “low response danger sign” for the adjustment module to use.

[0127] In addition, this module can automatically summarize the daily training data and output a status trend report for continuous training in chronological order.

[0128] IV. Intervention Module:

[0129] The intervention module is the core functional unit of the system. It is responsible for taking the assessment results and judgment status as input, calling upon the internal intervention resource library, and dynamically generating individualized intervention procedures. Its operation process includes:

[0130] 1) Intervention content generation: Based on the grouping of subjects, the appropriate Tai Chi training content is called from the training module library, including but not limited to the five-form Tai Chi, ten-form Tai Chi, and standard twenty-four-form Tai Chi, and the training time, beat, standing posture guidance, etc. are set in combination with the rhythm control parameters.

[0131] 2) Intervention strategy matching: The intervention plan is automatically generated by combining modules such as emotion regulation (e.g., mindfulness meditation, guided breathing), cognitive reconstruction (e.g., self-statement, feedback review), and behavioral incentive (e.g., stage badge rewards, task unlocking mechanism) from the strategy library.

[0132] 3) Intervention execution process support: The system provides multiple output methods such as voice prompts, visual feedback (such as action mirror projection), and training rhythm light strips to guide users to complete the actions;

[0133] 4) Collect feedback data during training and judge in real time whether the training effect meets the standard; those who do not meet the standard will be prompted to conduct the next round of training at the same difficulty, or switch to other rhythm modules.

[0134] V. Adjustment Module:

[0135] The adjustment module is responsible for dynamically adjusting the training path, strategy allocation, and content pace during the middle or late stages of the intervention to ensure maximum intervention effectiveness. This module has the following functions:

[0136] 1) Dynamic state analysis: By integrating the cognitive-emotional consistency matrix provided by the integrated module with the execution feedback of the intervention module, the state of each training unit is labeled (good / average / abnormal).

[0137] 2) Automatic adjustment mechanism: For an individual who is marked as "abnormal" twice in a row, the system will lower the intervention level by one level, that is, change the training content from 24 exercises to 10 exercises, or change the emotion regulation from mindfulness to guidance.

[0138] 3) Strategy fine-tuning function: For some trainees with slight execution deviations, the system prioritizes adjusting the feedback method (e.g., the frequency of voice prompts, the speed of rhythmic animation) without directly changing the training content, thus maintaining the continuity of intervention;

[0139] 4) Periodic assessment output: A training cycle report is generated every 7 days. The report includes cognitive execution trend curves, emotional fluctuation range analysis, training content adjustment trajectory, etc., providing visual reference for human intervention personnel.

[0140] Additional features design:

[0141] To ensure user engagement and autonomy in long-term training, the system also includes training calendar reminders, an incentive points system, personalized training suggestions, and stage goal setting functions. Users can view the completion percentage of each stage, earned badges, recommended exercises for the next stage, and psychological adjustment suggestions on the client.

[0142] System deployment architecture:

[0143] This system adopts a front-end and back-end separation architecture. The front-end is a cross-platform web and mobile application that integrates modules such as camera capture, image recognition, voice prompt playback, and interactive interface. The back-end is deployed on a cloud server and is responsible for data storage, recognition algorithm inference, intervention content scheduling, and model push updates. All data synchronization processes use encrypted communication protocols to ensure user privacy and security.

[0144] In summary, this system implementation achieves integrated functions such as deep fusion of cognitive state and motor performance, multimodal emotion recognition, and adaptive optimization of intervention strategies. It has the advantages of strong scalability, flexible response, and high user participation, and can provide solid technical support for cognitive rehabilitation training in multiple scenarios in the future.

[0145] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A Taijiquan cognitive intervention method based on sports psychology, characterized in that, Comprising the following steps: S1, collecting historical data of the intervention object, extracting cognitive state indicators, and preliminarily grouping the intervention object; S2, standardizing the cognitive state indicators of the intervention object, constructing a score vector, and calculating the cognitive state score by weighting method according to the set weight factor, and adjusting the grouping according to the score result; wherein the weighted items of the cognitive state score include: psychological test score, cognitive state score and behavior stability score; S3, obtaining the learning video of Taijiquan of the intervention object, combining the corresponding cognitive state score for content analysis, and judging whether the action execution state is in the standard range; S4, obtaining the emotion index of the intervention object, integrating and analyzing the emotion index and the cognitive state score according to the preset score weight and matching rule, and completing the final grouping; S5, developing a grouping intervention scheme, implementing Taijiquan training and psychological intervention, collecting and analyzing the execution process data in real time, evaluating the intervention completion degree, dynamically adjusting the intervention scheme, and reaching the preset intervention target.

2. The Taijiquan cognitive intervention method based on sports psychology according to claim 1, characterized in that: In step S1, the cognitive state indicators include concentration, understanding and memory, and the collection of historical data of the intervention object includes the following steps: Collecting the historical psychological test score, learning time, basic evaluation score, movement error times and emotional state fluctuation value of the intervention object; Standardizing the historical psychological test score, basic evaluation score and learning time respectively to obtain three groups of standardized basic vectors; The above basic vectors are set as a group, the movement error times and the emotional state fluctuation value are set as compensation vectors, and the cognitive state indicators are obtained by solving the weighted distance between the basic vectors and the compensation vectors; The calculation formula of the weighted distance is: ； wherein, , , is a base vector corresponding item, , , is a compensation vector corresponding item, , , is an association weight of each dimension and cognitive state score, and .

3. The Taijiquan cognitive intervention method based on sports psychology according to claim 2, characterized in that: In step S4, the acquisition of the emotion index includes the following steps: Collecting the facial expression image and physiological signal data of the intervention object, extracting the expression parameters from the grayscale and key feature point detection algorithm of the expression image; The physiological signal is processed by band pass filtering to obtain the heart rate variability value and the skin electric reaction intensity; According to the time synchronization mechanism, the expression parameters and the feature vectors of the physiological signals are paired to calculate their weighted average values; The signals in neutral state are identified and removed by setting a threshold, and the current emotion index is obtained by proportion calculation of the remaining signals.

4. The Taijiquan cognitive intervention method based on sports psychology according to claim 1, characterized in that: In step S5, developing a grouping intervention scheme includes the following steps: Establishing a training module library and an intervention strategy module library; According to the final grouping result, the corresponding training module and intervention strategy are extracted from the above module library to generate the intervention scheme; Wherein, the training module includes primary module, intermediate module and advanced module; the intervention strategy module includes cognitive guidance module, feedback reinforcement module and emotion regulation module.

5. The Taijiquan cognitive intervention method based on sports psychology according to claim 4, characterized in that: During the implementation of the intervention scheme, the execution process data is obtained after completing each group of training modules, the execution process data includes execution period, completion degree and feedback difference value, the feedback difference value is the difference value between the final similarity of the execution action and the standard similarity of the reference action.

6. The Taijiquan cognitive intervention method based on sports psychology according to claim 1, characterized in that: The execution process data is used in the process of calculating the intervention completion degree, the intervention completion degree is set as the objective function, the parameter is updated by using the iterative process based on Bayesian optimization, and the dynamic constraint is set for the execution period; the expression of the objective function is: ； ； wherein, is the maximum expected corresponding combination to be obtained in the Bayesian optimization iteration, is the optimization function is the expected output value of the optimization function is the mathematical expectation operator, is the execution period, is the average final similarity, is the feedback difference; is the maximum solving function, representing the combination of that maximizes the expected output value of the function , and are the sub-item intensity adjustment coefficients of the optimization function , which are set by empirical values, is the standard execution period.​ 7. A Tai Chi cognitive intervention system based on sports psychology, for performing the method of any one of claims 1 to 6, characterized in that, Comprising: The evaluation module is configured to collect historical data of the intervention object, extract cognitive state indicators, and perform preliminary grouping. The determination module is configured to obtain cognitive state scores and determine whether the action execution state of the intervention object is standardized. The integration module is configured to obtain an emotional index, establish a correlation model between emotion and cognitive state, and complete final grouping. The intervention module is configured to develop an intervention program and perform Taijiquan training and psychological intervention, and output intervention completion degree of each stage. The adjustment module is configured to dynamically adjust the intervention program according to the intervention completion degree until the intervention target is reached.

8. The Taijiquan cognitive intervention system based on sports psychology according to claim 7, characterized in that: The evaluation module is further configured to collect historical psychological test scores, learning duration, basic test scores, movement error times, and emotional fluctuation values, perform standardization processing, and construct a basic vector and a compensation vector, calculate a cognitive state indicator by using a weighted distance, and the calculation formula of the weighted distance is: ； wherein, , , is a base vector corresponding item, , , is a compensation vector corresponding item, , , is an association weight of each dimension with the cognitive state score, and .

9. The Taijiquan cognitive intervention system based on sports psychology according to claim 8, characterized in that: The feedback module is further configured to collect facial expressions and physiological signals of the intervention object, input the facial expressions and physiological signals into an emotional analysis algorithm, and output a corresponding emotional index. The determination module integrates the emotional index and the cognitive score to determine intervention consistency, and if the intervention consistency is not consistent, the intervention object is re-grouped and a warning signal is issued.

10. The Taijiquan cognitive intervention system based on sports psychology according to claim 7, characterized in that: When developing the intervention program, the intervention module calls a training module library and an intervention strategy module, selects corresponding modules according to the final grouping, and combines to generate multiple intervention programs for intervention execution and adjustment.

Images