Comprehensive evaluation method for equestrian training of children

Abstract

The invention relates to the technical field of children rehabilitation training and intelligent evaluation, and discloses a comprehensive evaluation method for children equestrian training, which comprises the following steps: acquiring basic parameters of user training data through a sensor, the basic parameters including training duration, movement speed, movement amplitude, turning angle, uphill and downhill amplitude and jumping height; performing data preprocessing on the basic parameters to generate evaluation parameters; inputting the evaluation parameters into an evaluation model constructed based on a gravity center balance algorithm for comprehensive calculation, and outputting an evaluation result including athletic ability, attention, emotion dimensions and balance keeping time proportion; and according to an evaluation result, allocating an adaptive comprehensive training item to the user, wherein the comprehensive training item comprises live-action video passive training, game active control training and interactive scene training. The exercise ability, attention, emotion and balance ability of the children can be scientifically evaluated, and it is ensured that the training effect better meets the rehabilitation requirements of the children.

Description

Technical Field

[0001] The present invention relates to the technical field of children's rehabilitation training and intelligence assessment, and in particular to a comprehensive assessment method for children's equestrian training. Background Art

[0002] Equestrian therapy is a rehabilitation method that uses the regular movement patterns of horses and all activities of human-horse interaction to conduct physical, occupational or speech therapy. It is an important part of the comprehensive and holistic rehabilitation training program implemented to achieve the ultimate functional rehabilitation goal.

[0003] Traditional children's equestrian training mostly relies on the coach's experience for evaluation and guidance. Although this method can provide certain feedback, it lacks objectivity and scientificity, and it is difficult to accurately grasp the children's training status and rehabilitation needs. In addition, the existing training equipment and evaluation system have failed to fully integrate advanced data collection and analysis methods, resulting in insufficient feedback during the training process and a lack of dynamic adjustment mechanisms. Therefore, the field of children's equestrian training urgently needs a more accurate, efficient and individually adaptable training evaluation system.

[0004] Existing children's equestrian training methods mainly focus on the cultivation of basic skills. However, these methods usually ignore the comprehensive evaluation of children's sports performance, especially in the comprehensive evaluation of multiple dimensions such as dynamic balance, athletic ability, attention, and emotions. Traditional training relies on the coach's observation and experience for adjustment, and cannot provide real-time feedback on training effects. It also has poor adaptability to individual differences in training. Therefore, the existing technology lacks an effective method that can comprehensively evaluate children's performance during training and adjust training programs based on the evaluation results.

[0005] At present, there are many problems with the existing children's equestrian training methods, including the lack of real-time and comprehensive evaluation methods, and it is difficult to effectively identify the impact of factors such as children's distraction and emotional fluctuations during training on the training effect; the technical solution provided by our invention can generate evaluation results in real time during children's training by comprehensively evaluating multiple dimensions such as athletic ability, emotions and attention, and combining dynamic training scene adjustments, providing a scientific basis for personalized training, and solving the subjectivity and limitations of traditional training methods. Summary of the invention

[0006] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract of the specification and the title of the invention of this application to avoid blurring the purpose of this section, the abstract of the specification and the title of the invention, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0007] In view of the above existing problems, the present invention is proposed.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions: Step 1, collecting basic parameters of user training data through sensors, wherein the basic parameters include training time, movement speed, movement amplitude, turning angle, uphill and downhill amplitude and jumping height;

[0009] Step 2: Preprocess the basic parameters to generate evaluation parameters;

[0010] Step 3: Input the evaluation parameters into the evaluation model constructed based on the center of gravity balance algorithm for comprehensive calculation, and output the evaluation results including the athletic ability, attention, emotional dimension and the proportion of time to maintain balance;

[0011] Step 4: Allocate an adaptive comprehensive training program for the user according to the evaluation result, wherein the comprehensive training program includes passive training of real-life video, active game control training, and interactive scene training.

[0012] As a preferred solution of the comprehensive evaluation method for children's equestrian training of the present invention, in the step 1, four sensor readings of the user's sitting position are obtained by a pressure sensor, and the center of gravity coordinates are calculated according to the sensor readings, which specifically includes:

[0013] A plane coordinate system based on the saddle force plate is established, and four pressure sensors are respectively located at four vertices of the saddle force plate;

[0014] Calculate the total pressure value of the four sensors:

[0015] F total =F A +F B +F C +F D

[0016] Among them, F total is the total pressure value of the four pressure sensors, F A 、F B 、F C 、F D The pressure readings of the four pressure sensors;

[0017] Based on the principle of moment balance, calculate the center of gravity coordinates (x cg ,y cg ), the formula is:

[0018]

[0019] Among them, x cg is the x coordinate of the center of gravity, y cgis the y coordinate of the center of gravity, x1, x2, x3, x4 are the x coordinates of the four pressure sensors, and y1, y2, y3, y4 are the y coordinates of the four sensors.

[0020] As a preferred embodiment of the comprehensive evaluation method for children's equestrian training of the present invention, the evaluation model is constructed by:

[0021] Preset the corresponding center of gravity stability range for different actions and establish an action-center of gravity range database;

[0022] The center of gravity coordinates acquired in real time are compared with the center of gravity stability range in the action-center of gravity range database to determine whether the user is in a balanced state.

[0023] As a preferred embodiment of the comprehensive evaluation method for children's equestrian training of the present invention, the evaluation of athletic ability includes static sitting stability evaluation and dynamic conversion stability evaluation, wherein:

[0024] When conducting a static sitting posture stability assessment, the center of gravity deviation data of the user performing a specified action is recorded when the riding machine is stationary. If the deviation is within the preset threshold, it is judged as qualified;

[0025] When conducting dynamic conversion stability assessment, the instantaneous change of the user's center of gravity offset is analyzed during the start-stop conversion process of the riding machine. If the change amplitude meets the preset standard, it is judged as qualified.

[0026] As a preferred embodiment of the comprehensive evaluation method for children's equestrian training of the present invention, the evaluation of attention includes:

[0027] The user's head posture data is collected through the camera, and the angle between the head direction and the preset positive direction is calculated;

[0028] If the proportion of time that the angle is within 15 degrees exceeds a set threshold, it is determined that the attention is focused.

[0029] As a preferred embodiment of the comprehensive evaluation method for children's equestrian training of the present invention, the emotion evaluation includes:

[0030] Use facial expression recognition technology to extract the user's expression features during training;

[0031] The features are matched with a preset emotion classification model and an emotion state score is output.

[0032] As a preferred solution of the comprehensive evaluation method for children's equestrian training of the present invention, the allocation of the comprehensive training items includes:

[0033] Filter training scenarios that match the user's capabilities based on the evaluation results, where the training scenarios include at least basic scenarios, advanced scenarios, and challenge scenarios;

[0034] In interactive scene training, the game difficulty parameters are dynamically adjusted according to the stability of the user's center of gravity. The game difficulty parameters include movement speed, obstacle density and task complexity.

[0035] As a preferred embodiment of the comprehensive evaluation method for children's equestrian training of the present invention, the interactive scene training includes three types: riding adventure scene, horse racing scene and traffic rules scene, wherein:

[0036] If it is the riding adventure scene, the user controls the movement of the virtual horse by adjusting the center of gravity, and completes the interactive tasks of pulling the reins to turn and waving hands;

[0037] If it is the horse racing scene, the user needs to maintain a stable center of gravity to increase the speed of the virtual horse and adjust the posture according to the real-time road conditions;

[0038] If it is the traffic rules scenario, the user needs to observe the virtual traffic lights and control the horse to comply with traffic instructions.

[0039] As a preferred embodiment of the comprehensive evaluation method for children's equestrian training of the present invention, the method further comprises a spinal posture evaluation, wherein:

[0040] Obtain the three-dimensional coordinates of the user's spinal bones through a bone tracking device;

[0041] The angle between the spine vector and the vertical ground direction is calculated, and if the angle exceeds a preset threshold, a posture correction suggestion is generated.

[0042] Beneficial effects of the present invention:

[0043] 1. Through accurate data collection, accurate input is provided for subsequent steps, ensuring the objectivity and accuracy of the evaluation, and laying the foundation for the allocation of customized training programs;

[0044] 2. Data preprocessing optimizes the quality of input data, improves the accuracy of model calculations, and provides clearer and more stable data support for subsequent comprehensive evaluations, thereby enhancing the accuracy and reliability of evaluation results;

[0045] 3. Through a comprehensive multi-dimensional evaluation model, the user's training performance is fully and accurately understood. It not only provides feedback on athletic ability, but also takes into account psychological factors such as children's emotions and attention during training. This comprehensive evaluation method helps to identify the user's deficiencies and potential and further optimize the training effect.

[0046] 4. Through personalized training project allocation, the training intensity and difficulty can be adjusted according to the user's actual evaluation results, which not only avoids the monotony of training and the risk of overtraining, but also enhances the fun of training and user participation, improves training effects, enhances children's long-term training interest, and ensures better compliance with children's rehabilitation needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0048] Figure 1 It is a flow chart of the comprehensive evaluation method for children's equestrian training shown in the present invention. DETAILED DESCRIPTION

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0050] Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without making any creative work should fall within the scope of protection of the present invention.

[0051] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0052] According to an embodiment of the present invention, Figure 1 The flowchart shown is a comprehensive assessment method for children's equestrian training, including:

[0053] S1. Collect basic parameters of user training data through sensors, including training duration, movement speed, movement range, turning angle, uphill and downhill range, and jumping height. Among them, what needs to be explained in this step is:

[0054] High-precision pressure sensors are installed at the four vertices (A, B, C, D) of the saddle force plate to monitor the pressure value (F A 、F B 、F C 、F D );

[0055] A six-axis inertial measurement unit (IMU) (including accelerometer and gyroscope) is integrated at the bottom of the saddle to collect movement speed, movement amplitude, turning angle and jumping height;

[0056] Deploy the altimeter / tilt sensor on the horseback stand to calculate the uphill and downhill slopes by measuring the inclination angle;

[0057] The timing module is embedded in the system controller to record the total training time and the duration of each segmented action.

[0058] It should be noted that the basic parameters are acquired synchronously, including:

[0059] The IMU's accelerometer can collect the user's acceleration data during training in real time, and calculate the user's instantaneous movement speed by combining the gyroscope's angular velocity data and compensating for centrifugal force errors;

[0060] The displacement trajectory of the IMU obtains the amplitude of the movement by integrating the acceleration data. The system detects the user's movement amplitude based on dynamic changes. If a sudden change in amplitude occurs during the calculation process, it is compared with the pressure sensor data to verify whether the abnormal amplitude is caused by imbalance, thereby improving the accuracy of the data.

[0061] The turning angle is calculated by the yaw angle change recorded by the gyroscope and correlated with the lateral offset of the center of gravity of the pressure sensor for verification;

[0062] The up and down slope amplitude is measured by the altimeter to measure the vertical displacement change, and the pitch angle of the tilt sensor is combined to verify the slope consistency;

[0063] The jump height is calculated by quadratically integrating the vertical acceleration data of the IMU and by using the instantaneous release characteristics of the saddle pressure (F total approaches zero) triggers the jump event marker.

[0064] In an optional embodiment, a plane coordinate system is established with the saddle force plate as a reference, and the coordinates of the four pressure sensors are defined as: A (x1, y1), B (x2, y2), C (x3, y3), and D (x4, y4). When the system is initialized, the sensor zero point calibration is performed to eliminate the influence of environmental noise and the saddle's own weight.

[0065] It should be further explained that, in this embodiment, the center of gravity coordinates are used to dynamically reflect the user's sitting balance state, providing a reference for parameters such as movement amplitude and turning angle.

[0066] As an example, four sensor readings of the user's sitting position are obtained through a pressure sensor, and the center of gravity coordinates are calculated based on the sensor readings, including:

[0067] A plane coordinate system based on the saddle force plate is established, and four pressure sensors are located at the four vertices of the saddle force plate;

[0068] Calculate the total pressure value of the four sensors:

[0069] F total =F A +F B +F C +F D

[0070] Among them, F total is the total pressure value of the four pressure sensors, F A 、F B 、F C 、F D The pressure readings of the four pressure sensors;

[0071] Based on the principle of moment balance, calculate the center of gravity coordinates (x cg ,y cg ), the formula is:

[0072]

[0073] Among them, x cg is the x coordinate of the center of gravity, y cg is the y coordinate of the center of gravity, x1, x2, x3, x4 are the x coordinates of the four pressure sensors, and y1, y2, y3, y4 are the y coordinates of the four sensors.

[0074] Preferably, by collecting basic parameters of the training process through sensors, such as training duration, movement speed, movement amplitude, turning angle, uphill and downhill amplitude, and jumping height, various dynamic indicators of the user in training can be monitored in real time to ensure accurate quantitative evaluation of the training situation, avoiding the limitations of traditional training methods that rely on subjective judgment and experience. The precise data provided by the sensor provides an accurate basis for subsequent analysis and model evaluation, ensuring the reliability of the evaluation results.

[0075] S2. Perform data preprocessing on basic parameters to generate evaluation parameters. The following should be noted in this step:

[0076] For each sensor's measurement data, set a value range. If the detection data exceeds the set range, it is considered abnormal and removed. For example:

[0077] The pressure sensor reading range is set to F A ,F B ,F C ,F D ∈[0,150]kg;

[0078] The range of motion speed is set to v∈[0,10]m / s;

[0079] The jump height range is set to h jump ≤1.5m.

[0080] For example, if there is a contradiction between the data, it will be deemed invalid; for example:

[0081] If the jump height h jump >0 but the total pressure sensor reading F total ≠0, it is considered as invalid data and the data segment is discarded.

[0082] Furthermore, for short-term missing data (missing duration ≤ 200ms), linear interpolation was used to fill the data; for long-term missing data (missing duration > 200ms), they were directly truncated and recorded as invalid intervals.

[0083] In an optional implementation, a sliding window mean filtering method is used to process the pressure sensor data, the window size is set to 100ms (ie, 10 sampling points), and the smoothing formula is:

[0084]

[0085] in, represents the i-th smoothed pressure sensor data, N = 10, which is the number of sampling points;

[0086] For acceleration and angular velocity data (i.e., IMU dynamic parameters), Kalman filtering is applied to suppress high-frequency noise;

[0087] Furthermore, in order to remove the IMU data jitter caused by the saddle vibration, the wavelet transform denoising algorithm was used, and the effective frequency band signal (0.1-20 Hz) was extracted.

[0088] Normalize all parameters to the interval [0,1], and use the timestamp matching algorithm to ensure that the synchronization error of multiple sensor data is ≤10ms;

[0089] Furthermore, the standard deviation σ of the barycentric coordinates is calculated x ,σ y , to reflect the sitting balance, extract the envelope area of ​​the center of gravity trajectory, and the calculation formula is:

[0090]

[0091] Among them, S cg is the envelope area of ​​the center of gravity trajectory, (x i ,y i ) is a point on the trajectory;

[0092] Then extract the acceleration mutation point from the motion speed v. If the acceleration mutation rate It is marked as an emergency stop or start event;

[0093] For example, for the turning angle θ turn Calculate the smoothness index, the calculation formula is:

[0094]

[0095] Among them, R smooth It is a smoothness index, which indicates the smoothness of the change of turning angle;

[0096] Use weighted fusion algorithm to generate comprehensive balance coefficient K balance , the formula is:

[0097]

[0098] Among them, α and β are weights, and α+β=1;

[0099] The final output evaluation parameters are displayed in the form of a set:

[0100] Params eval ={σ x ,σ y ,S cg ,K balance ,R smooth ,N jump ,T stable}

[0101] Among them, σ x ,σ y is the standard deviation of the barycentric coordinates, S cg is the envelope area of ​​the center of gravity trajectory, K balance is the comprehensive balance coefficient, R smooth is the smoothness index of the turning angle, N jump is the number of jumps, T stable It is to balance the duration ratio.

[0102] As an example, the data package outputs parameters in segments according to the time window (default 5s) in the form of a JSON array:

[0103]

[0104] It should be noted that the validity and accuracy of the data are ensured by filtering out noise, removing abnormal data, and smoothing dynamic parameters. Especially in the dynamic training process, the sliding window algorithm can remove instantaneous noise and make the evaluation parameters more in line with the actual training status. This not only improves the quality of the data, but also enhances the system's adaptability to different training environments and user performance.

[0105] S3, input the evaluation parameters into the evaluation model constructed based on the center of gravity balance algorithm for comprehensive calculation, and output the evaluation results including the athletic ability, attention, emotional dimension and the proportion of time to maintain balance. Among them, what needs to be explained in this step is:

[0106] In an optional embodiment, the evaluation model is constructed in a manner including:

[0107] Preset the corresponding center of gravity stability range for different actions and establish an action-center of gravity range database;

[0108] The center of gravity coordinates obtained in real time are compared with the center of gravity stability range in the action-center of gravity range database to determine whether the user is in a balanced state.

[0109] As an example, the establishment of the action-center of gravity range database includes:

[0110] Collect the center of gravity coordinate range of standard users in different actions (such as flat riding, turning, jumping), and collect at least 100 sets of sample data for each action;

[0111] Perform statistical analysis on the center of gravity coordinates of each action and set the dynamic balance interval [x min ,x max ]×[y min ,y max ] and store it in the database, for example:

[0112] Flat riding: x∈[-30,30]mm,y∈[-20,20]mm;

[0113] Turning action: According to the turning angle θ turn Dynamically adjust the x range, the formula is:

[0114] x threshold = ±(50+0.5·|θ turn |

[0115] Database structure: Action types, balance intervals, and associated parameters (slope, jump height, etc.) are stored in JSON format, for example:

[0116]

[0117] Furthermore, input the real-time center of gravity coordinates (x real ,y real ), query the balance interval corresponding to the current action, if x real ∈[x min ,x max ] and y real ∈[y min ,ymax ], it is judged as a balanced state; otherwise it is marked as unbalanced.

[0118] In an optional embodiment, the assessment of movement ability includes static sitting stability assessment and dynamic transition stability assessment, wherein:

[0119] When conducting a static sitting posture stability assessment, the center of gravity deviation data of the user performing a specified action is recorded when the riding machine is stationary. If the deviation is within the preset threshold, it is judged as qualified;

[0120] When conducting dynamic conversion stability assessment, the instantaneous change of the user's center of gravity offset is analyzed during the start-stop conversion process of the riding machine. If the change amplitude meets the preset standard, it is judged as qualified.

[0121] As an example, a static sitting stability assessment includes:

[0122] When the riding machine is stationary, record the center of gravity deviation data when the user maintains a sitting position {x i ,y i}(sampling interval 0.2 seconds);

[0123] Calculate the standard deviation σ of the center of gravity shift static :

[0124]

[0125] in, and are the means of the center of gravity offset data respectively;

[0126] If σ static If the diameter is ≤15mm, it is considered qualified; otherwise, it is recommended to conduct core muscle strengthening training.

[0127] As an example, dynamic conversion stability assessment includes:

[0128] During the start-stop transition of the riding machine (from 1 second before to 2 seconds after), record the center of gravity deviation rate v cg :

[0129]

[0130] Where Δx and Δty are the displacements of the center of gravity in the x and y directions, and Δt is the time interval;

[0131] If the maximum instantaneous rate It is judged as qualified.

[0132] In an optional embodiment, the assessment of attention includes:

[0133] The user's head posture data is collected through the camera, and the angle between the head direction and the preset positive direction is calculated;

[0134] If the proportion of time that the angle is within 15 degrees exceeds the set threshold, it is determined that the attention is focused.

[0135] As an example, the 3D coordinates of the key points of the user's head bones are obtained through the Kinect camera. Calculate the head heading vector;

[0136] Define the positive direction as the direction of the camera optical axis

[0137] Calculate the angle between the head orientation and the camera optical axis:

[0138]

[0139] The statistical angle is less than or equal to 15 degrees. attention :

[0140]

[0141] If P attention ≥80%, it is judged as concentration.

[0142] In an optional embodiment, the assessment of emotions includes:

[0143] Use facial expression recognition technology to extract the user's expression features during training;

[0144] Match the features with the preset emotion classification model and output the emotion state score.

[0145] As an example, the OpenFace library is used to extract 52 facial action unit (AU) intensity values ​​to form a feature vector

[0146] Using the pre-trained support vector machine (SVM) classifier, the feature vector Input, output emotion label (such as excitement, tension, calm);

[0147] The sentiment scoring formula is defined as:

[0148]

[0149] Among them, w i is the emotion weight (e.g., excitement = 0.4, tension = -0.3, calmness = 0.2), p i is the classification probability.

[0150] In an optional embodiment, the evaluation of the proportion of time to maintain equilibrium includes:

[0151] With time t as the horizontal axis, the lateral displacement of the center of gravity x cg The vertical axis is sampled every 0.2 seconds, and a continuous line graph is drawn;

[0152] Statisticsx cg ∈[x min ,x max ] time points N stable , total time points N total :

[0153]

[0154] The mathematical expression formula of the balance time ratio is:

[0155]

[0156] Among them, R balance To balance the time ratio.

[0157] For example, the comprehensive score S is calculated based on different evaluation results. total :

[0158] S total =0.35·S static +0.25·S dynamic +0.2·P attention +0.1·S emotion +0.1·R balance

[0159] in, is the normalized score for static stability,

[0160] Standardized scores for dynamic stability;

[0161] If the comprehensive score S total A high score means that the user performs well in static stability, dynamic stability, attention, emotional control, and the ability to maintain balance;

[0162] If the comprehensive score S total A low score may indicate that the user has significant room for improvement in a certain area (such as dynamic stability or attention).

[0163] Exemplarily, the evaluation results are output in a structured format, such as:

[0164]

[0165] Preferably, this step performs comprehensive calculations by inputting the preprocessed evaluation parameters into an evaluation model constructed based on the center of gravity balance algorithm, which can comprehensively evaluate multiple dimensions such as athletic ability, attention, and emotions; the center of gravity balance algorithm can accurately determine whether the user maintains balance during training. The model not only considers static balance, but can also evaluate dynamic balance, and combines the user's performance in exercise to form accurate evaluation results.

[0166] S4. Allocate an adapted comprehensive training project to the user based on the evaluation results. The comprehensive training project includes passive training of real-life video, active game control training, and interactive scene training. It should be noted that the allocation of comprehensive training projects includes:

[0167] Filter training scenarios that match user capabilities based on the evaluation results. The training scenarios include at least basic scenarios, advanced scenarios, and challenge scenarios.

[0168] In interactive scene training, the game difficulty parameters are dynamically adjusted based on the user's center of gravity stability. The game difficulty parameters include movement speed, obstacle density and task complexity.

[0169] As an example, the basic scenario applies to the comprehensive score S total For ≤50% of users, the training content includes static balance exercises (such as sitting posture maintenance) and low-complexity command responses (such as slow straight-line riding);

[0170] As an example, the advanced scenario applies to 50% total For ≤80% of users, the training content includes dynamic balance challenges (such as riding on a gentle slope and turning at a small angle) and multi-task coordination (such as waving while riding);

[0171] As an example, the challenge scenario applies to S total >80% of users, training content includes high-speed turns, continuous jumps and complex environmental responses (such as avoiding obstacles);

[0172] Exemplarily, the movement speed adjustment rule is:

[0173] According to the center of gravity stability coefficient K balance Dynamically adjust the speed of the virtual horse v game :

[0174] v game =v base ·(1+0.5·K balance )

[0175] Among them, v base is the base speed of the scene. For example, in the base scene, v base =3m / s, and K balance is the balance coefficient;​

[0176] Exemplarily, the adjustment rule of obstacle density is:

[0177] If the user's balance time ratio R balance ≥70%, the obstacle generation interval is shortened by 20%;

[0178] Exemplarily, the adjustment rule of task complexity is:

[0179] According to the attention score P attention , dynamically increase parallel tasks:

[0180]

[0181] in, It means rounding down, which means the number of tasks is adjusted according to the attention score.

[0182] As an example, interactive scenario training includes three types: cycling adventure scenario, horse racing scenario, and traffic rules scenario. Among them:

[0183] If it is a riding adventure scene, the user controls the movement of the virtual horse by adjusting the center of gravity, and completes interactive tasks such as pulling the reins to turn and waving hands;

[0184] For example, the lateral displacement of the user's center of gravity x cg Mapped to the virtual horse's steering angle θ turn :

[0185] θ turn = k·x cg ,k=0.1° / mm

[0186] Where k is the conversion coefficient, which represents the relationship between the center of gravity offset and the steering angle.

[0187] When the user's hand wave is detected (using Kinect skeleton to track wrist acceleration a wrist >15m / s 2 ), triggering the NPC dialogue event.

[0188] If it is a horse racing scene, the user needs to maintain a stable center of gravity to increase the speed of the virtual horse and adjust the posture according to the real-time road conditions;

[0189] For example, if the standard deviation of the centroid for 5 consecutive seconds is x ≤10mm, the speed of the virtual horse increases by 20%; by randomly generating the slope slope ∈[-15°,15°], the user needs to adjust the sitting posture to maintain balance.

[0190] If it is a traffic rules scene, the user needs to observe the virtual traffic lights and control the horse to follow traffic instructions.

[0191] For example, the virtual traffic light switches state every 30 seconds, and the user needs to move the center of gravity backward (y cg <-20mm) to brake; if you fail to brake in time, points will be deducted.

[0192] In an optional embodiment, the method provided in this embodiment further includes spinal posture assessment, wherein:

[0193] Obtain the three-dimensional coordinates of the user's spinal bones through a bone tracking device;

[0194] The angle between the spine vector and the vertical ground direction is calculated. If the angle exceeds the preset threshold, a posture correction suggestion is generated.

[0195] As an example, the coordinates P of the upper end of the spine are obtained through Kinect skeleton tracking. top (x1, y1, z1) and the lower coordinate P bot (x2,y2,z2);

[0196] Spine vector definition:

[0197]

[0198] Vertical reference vector:

[0199]

[0200] Angle calculation formula:

[0201]

[0202] Among them, θ spine is the angle between the spine and the vertical direction;

[0203] If θ spine >10° for more than 5 seconds, the system will generate a voice prompt "Please straighten your back". After more than 3 warnings, the game difficulty will be automatically reduced by one level (such as switching from the challenge scene to the advanced scene).

[0204] In an optional embodiment, the training system displays the current training scene, real-time score and spinal posture monitoring diagram (dynamic angle curve between the spinal vector and the vertical line), and the balance line graph superimposes the recommended correction interval (such as θ spine ≤10° green area);

[0205] After each day's training, a PDF report is generated, which includes the trend graph of scores in each dimension, the percentage of abnormal spinal posture time, and training suggestions for the next stage (such as "need to increase core muscle training by 20%");

[0206] Exemplarily, the calculation formula for the abnormal spinal posture time ratio is:

[0207]

[0208] Among them, T warning is the time of abnormal spinal posture, T total is the total training time.

[0209] Preferably, according to the evaluation results, a comprehensive training program suitable for the user is assigned, including passive training with real-life videos, active control training with games, and interactive scene training. The training content can be tailored to each child's unique performance. This dynamic adjustment mechanism ensures that each child receives appropriate training tasks based on his or her motor ability, attention, and emotional state, avoiding the effect bottleneck caused by homogenized training.

[0210] At the same time, by dynamically adjusting the game difficulty parameters in interactive scenarios, the challenge of training is further increased, making the training process more personalized and interesting, and promoting children's continuous learning and progress.

[0211] It should also be noted that compared with the existing technology, our invention proposes a training method based on a multi-dimensional evaluation model, which collects children's training data in real time through sensors and combines it with the center of gravity balance algorithm for comprehensive calculation. It can scientifically evaluate children's motor skills, attention, emotions and balance abilities, thereby providing data support for training and ensuring that the training effect is more in line with children's rehabilitation needs.

[0212] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A comprehensive evaluation method for children's equestrian training, characterized in that: include: Step 1: Collect basic parameters of user training data through sensors, including training duration, movement speed, movement amplitude, turning angle, uphill and downhill amplitude, and jumping height; Step 2: Preprocess the basic parameters to generate evaluation parameters; Step 3: Input the evaluation parameters into the evaluation model constructed based on the center of gravity balance algorithm for comprehensive calculation, and output the evaluation results including the athletic ability, attention, emotional dimension and the proportion of time to maintain balance; Step 4: Allocate an adaptive comprehensive training program for the user according to the evaluation result, wherein the comprehensive training program includes passive training of real-life video, active game control training, and interactive scene training.

2. The comprehensive evaluation method for children's equestrian training according to claim 1, characterized in that: In the step 1, four sensor readings of the user's sitting position are obtained through a pressure sensor, and the center of gravity coordinates are calculated according to the sensor readings, which specifically includes: A plane coordinate system based on the saddle force plate is established, and four pressure sensors are respectively located at four vertices of the saddle force plate; Calculate the total pressure value of the four sensors: F total =F A +F B +F C +F D Among them, F total is the total pressure value of the four pressure sensors, F A 、F B 、F C 、F D The pressure readings of the four pressure sensors; Based on the principle of moment balance, calculate the center of gravity coordinates (x cg ,y cg ), the formula is: Among them, x cg is the x coordinate of the center of gravity, y cg is the y coordinate of the center of gravity, x1, x2, x3, x4 are the x coordinates of the four pressure sensors, and y1, y2, y3, y4 are the y coordinates of the four sensors.

3. The comprehensive evaluation method for children's equestrian training according to claim 1, characterized in that: The evaluation model is constructed in a manner including: Preset the corresponding center of gravity stability range for different actions and establish an action-center of gravity range database; The center of gravity coordinates acquired in real time are compared with the center of gravity stability range in the action-center of gravity range database to determine whether the user is in a balanced state.

4. The comprehensive evaluation method for children's equestrian training according to claim 1, characterized in that: The assessment of movement ability includes static sitting stability assessment and dynamic transition stability assessment, wherein: When conducting a static sitting posture stability assessment, the center of gravity deviation data of the user performing a specified action is recorded when the riding machine is stationary. If the deviation is within the preset threshold, it is judged as qualified; When conducting dynamic conversion stability assessment, the instantaneous change of the user's center of gravity offset is analyzed during the start-stop conversion process of the riding machine. If the change amplitude meets the preset standard, it is judged as qualified.

5. The comprehensive evaluation method for children's equestrian training according to claim 1, characterized in that: The attention assessment includes: The user's head posture data is collected through the camera, and the angle between the head direction and the preset positive direction is calculated; If the proportion of time that the angle is within 15 degrees exceeds a set threshold, it is determined that the attention is focused.

6. The comprehensive evaluation method for children's equestrian training according to claim 1, characterized in that: Assessment of emotions, including: Use facial expression recognition technology to extract the user's expression features during training; The features are matched with a preset emotion classification model, and an emotion state score is output.

7. The comprehensive evaluation method for children's equestrian training according to claim 1, characterized in that: The distribution of the comprehensive training program includes: Filter training scenarios that match the user's capabilities based on the evaluation results, where the training scenarios include at least basic scenarios, advanced scenarios, and challenge scenarios; In interactive scene training, the game difficulty parameters are dynamically adjusted according to the stability of the user's center of gravity. The game difficulty parameters include movement speed, obstacle density and task complexity.

8. The comprehensive evaluation method for children's equestrian training according to claim 7, characterized in that: The interactive scene training includes three types: cycling adventure scene, horse racing scene and traffic rules scene, among which: If it is the riding adventure scene, the user controls the movement of the virtual horse by adjusting the center of gravity, and completes the interactive tasks of pulling the reins to turn and waving hands; If it is the horse racing scene, the user needs to maintain a stable center of gravity to increase the speed of the virtual horse and adjust the posture according to the real-time road conditions; If it is the traffic rules scenario, the user needs to observe the virtual traffic lights and control the horse to comply with traffic instructions.

9. The comprehensive evaluation method for children's equestrian training according to claim 1, characterized in that: The method also includes a spinal posture assessment, wherein: Obtain the three-dimensional coordinates of the user's spinal bones through a bone tracking device; The angle between the spine vector and the vertical ground direction is calculated, and if the angle exceeds a preset threshold, a posture correction suggestion is generated.

Images