Whole body joint rehabilitation training device and control method thereof

Abstract

The invention relates to the technical field of rehabilitation instruments, in particular to a whole-body joint rehabilitation training device and a control method thereof. The method comprises the following steps: acquiring joint angles, IMU data and electromyographic signals when a user executes a plurality of standard actions; according to the joint angle, determining a joint motion range risk level of the target joint; according to the joint angle, the IMU data and the electromyographic signal, determining a compensatory behavior risk level and a cooperative efficiency score of a target joint; determining a training mode according to the risk level of the compensatory behavior, wherein the training mode at least comprises a general mode and a mode for the compensatory behavior; determining training parameters according to the joint motion range risk level, wherein the training parameters comprise resistance; adjusting the training parameters according to the collaborative efficiency score; and outputting the training mode and the adjusted training parameters. According to the method, the rehabilitation efficiency can be improved, and compensatory behaviors during training are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of rehabilitation equipment, and more particularly to a whole-body joint rehabilitation training device and a control method thereof. Background Art

[0002] Joint mobility disorder refers to a limited or abnormal range of motion in a joint due to various reasons, such as illness or trauma, which can severely impact a patient's limb function and quality of life. Existing range of motion training programs often use a standardized approach and fail to adapt to the user's specific needs. Furthermore, some patients are prone to compensatory behaviors during training, using abnormal movements of other joints or muscles to complete the training. This not only fails to effectively improve the range of motion of the target joint, but can also lead to new muscle strains or joint injuries.

[0003] Therefore, how to conduct effective joint mobility training and avoid compensatory behavior is a technical problem that needs to be solved urgently. Summary of the Invention

[0004] In order to solve the above-mentioned technical problems of not being able to obtain effective training and possibly causing compensatory behavior during joint range of motion training, the present invention provides solutions in the following aspects.

[0005] In a first aspect, the present invention provides a control method for a whole-body joint rehabilitation training device, comprising: collecting joint angles, IMU data and electromyographic signals when a user performs multiple standard movements; determining the joint mobility risk level of a target joint based on the joint angle; determining the compensatory behavior risk level and the synergy efficiency score of the target joint based on the joint angle, the IMU data and the electromyographic signal, the compensatory behavior risk level characterizing the user's tendency to utilize compensatory behavior, and the synergy efficiency score characterizing the synergy between the target joint and other joints; determining a training mode based on the compensatory behavior risk level, the training mode including at least a general mode and a mode for compensatory behavior; determining training parameters based on the joint mobility risk level, the training parameters including resistance; adjusting the training parameters based on the synergy efficiency score, the synergy efficiency score being positively correlated with the training parameters; outputting the training mode and the adjusted training parameters.

[0006] Furthermore, determining the training mode according to the compensatory behavior risk level includes: determining it according to a preset mapping relationship between the compensatory behavior risk level and the training mode.

[0007] Furthermore, training parameters are determined according to the joint mobility risk level, including: judging whether the joint mobility risk level is a level one risk; if so, determining the resistance to be provided according to the user's weight, and the resistance is positively correlated with the weight; if not, determining the resistance to be provided according to the current joint angle of the target joint, and the resistance is positively correlated with the current joint angle.

[0008] Furthermore, the resistance required is determined based on the current range of motion of the joint, including:

[0009]

[0010] Where F is the resistance to be provided, k is the reference coefficient, T is the time constant, t is the current training time, ROM curent is the current joint angle.

[0011] Furthermore, adjusting the training parameters according to the synergistic efficiency score includes: adjusting the benchmark coefficient according to the synergistic efficiency score to adjust the resistance, and the benchmark coefficient is positively correlated with the synergistic efficiency score.

[0012] Furthermore, during the training process, the method further includes: adjusting the training parameters at set time intervals.

[0013] Furthermore, based on the joint angle, the IMU data and the electromyographic signal, the compensatory behavior risk level and the synergistic efficiency score of the target joint are determined, including: fusing the joint angle and the IMU data to obtain a joint motion trajectory matrix; obtaining a muscle activation timing diagram based on the electromyographic signal; inputting the joint motion trajectory matrix and the muscle activation timing diagram into a trained evaluation model to obtain the synergistic efficiency score and the compensatory behavior risk level; the evaluation model is a machine learning model.

[0014] Furthermore, the joint mobility risk level of the target joint is determined based on the joint angle, including: matching the standard mobility range according to the target joint; and determining the joint mobility risk level based on the difference between the joint angle range of the user's target joint and the standard mobility range.

[0015] Furthermore, it also includes: displaying a virtual training scene through AR technology and outputting corresponding prompts.

[0016] In a second aspect, a whole-body joint rehabilitation training device is provided, comprising a flexible sensor, an IMU array, a high-density surface electromyography, wherein the flexible sensor is used to collect joint angles, the IMU array is used to collect IMU data, and the high-density surface electromyography is used to collect electromyography signals; an exoskeleton support used to provide the required resistance and assistance for training; and a controller connected with the flexible sensor, the IMU array, the high-density surface electromyography, and the exoskeleton support, used to realize the control method of the whole-body joint rehabilitation training device in the first aspect.

[0017] The method of the present application can provide corresponding training parameters according to the actual situation of the user, assist the user to effectively improve the joint range of motion training, thereby improving the efficiency of the user's rehabilitation; by adjusting the training mode and training parameters of the user according to the compensation behavior risk level and the synergy efficiency score of the user, the occurrence of compensation behavior can be reduced, thereby ensuring the rehabilitation efficiency and safety of the user training; real-time error correction is performed through AR technology, further improving the rehabilitation efficiency of the training. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a flowchart schematically showing a control method of a whole-body joint rehabilitation training device according to an embodiment of the present application;

[0019] Figure 2 is a structural block diagram schematically showing a whole-body joint rehabilitation training device according to an embodiment of the present application;

[0020] Figure 3 is a schematic diagram of an exoskeleton support in the prior art. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0022] The specific embodiments of the present application will be described in detail below with reference to the drawings.

[0023] Figure 1 is a flowchart schematically showing a control method of a whole-body joint rehabilitation training device according to an embodiment of the present application.

[0024] In a first aspect, the present application provides a control method of a whole-body joint rehabilitation training device, as Figure 1 shown, the control method of the present application comprises S1 to S6.

[0025] S1. Collect joint angles, IMU data, and electromyographic signals when the user performs multiple standard actions.

[0026] In this embodiment, IMU data includes angular velocity and acceleration. Flexible sensors are used to collect the joint angles of the body's major joints; an IMU (inertial measurement unit) array (integrated with a gyroscope and accelerometer) collects angular velocity and acceleration; and HD-sEMG electrodes (high-density surface electromyography) collect the electromyographic signals (i.e., muscle activation strength) of target muscle groups.

[0027] In one embodiment, the major joints of the human body may include: 14 joints: cervical vertebrae, thoracic vertebrae, lumbar vertebrae, sacroiliac joints, shoulder joints, elbow joints, wrist joints, radioulnar joints, metacarpophalangeal joints, fingertip joints, hip joints, knee joints, ankle joints and subtalar joints.

[0028] The user is prompted to perform multiple preset standard movements (e.g., standing upright, squatting, arm raising, etc.), each of which is maintained for 10 seconds. Joint angles, IMU data, and myoelectric signals are collected during this time. In one embodiment, Kalman filtering can also be used to eliminate baseline drift of sensors (including flexible sensors, IMU arrays, etc.), and then the user's motion baseline (the joint angle range of each joint) is established.

[0029] S2. Determine the joint mobility risk level, compensatory behavior risk level, and synergistic efficiency score of the target joint.

[0030] The joint mobility risk level represents the degree of difference between the user's joint mobility and normal joint mobility; the compensatory behavior risk level represents the user's tendency to use compensatory behavior; and the synergy efficiency score represents the synergy between the target joint and other joints.

[0031] For any of the 14 major joints, the joint mobility risk level can be determined based on the joint angle. Specifically, the normal range of motion (i.e., standard range of motion) corresponding to the joint is searched from an international standard database (containing the normal range of motion corresponding to each joint). The corresponding joint mobility risk level is then matched based on the difference between the joint angle range (composed of the minimum and maximum values ​​of the joint angle collected) and the standard range of motion corresponding to the joint.

[0032] In one embodiment, the span value of the joint angle range (maximum value minus minimum value) is divided by the span value of the corresponding standard range of motion to obtain a risk score, and then the corresponding joint range of motion risk level is matched according to the risk score. Specifically, when the risk score is less than 0.15, it is a level 1 risk; when the risk score is greater than or equal to 0.15 and less than 0.3, it is a level 2 risk; when the risk score is greater than or equal to 0.3 and less than 0.6, it is a level 3 risk; and when the risk score is greater than 0.6, it is a level 4 risk. In this way, the joint range of motion risk levels of the 14 major joints are obtained.

[0033] For example, if the joint angle range of the user's shoulder external rotation is 65-80°, and the standard range of motion of the shoulder external rotation found in the international standard database is 60-90°, the risk score of the user's shoulder joint is 0.5 (15 / 30=0.5), and the joint motion risk level is level three risk.

[0034] In one embodiment, a muscle activation heat map (e.g., quadriceps activation intensity between 0-100%) can be generated based on the electromyographic signal. The muscle activation heat map is used to display the activation intensity and timing characteristics of a specific muscle group during exercise. In one embodiment, a muscle activation heat map is generated based on the electromyographic signal and the corresponding time. In this embodiment, the mathematical expression of the muscle activation heat map H is:

[0035] H∈R M×T ;

[0036] Where M is the number of target muscle groups, T is the time step (corresponding to the duration of the movement), R is a set of real numbers, H is the muscle activation timing matrix (M rows and T columns), and H ij is an element in H, representing the activation intensity of the i-th muscle at the j-th time point (normalized to the interval [0,1]).

[0037] In one embodiment, the acceleration, angular velocity, and joint angle are fused (aligned by timestamps, such as synchronization via a precision time protocol) to generate a joint motion trajectory matrix (a core data structure describing the motion state of a multi-joint system, with 14 rows and 3 columns, in units of angle / acceleration). Specifically, the mathematical expression of the joint motion trajectory matrix P is:

[0038] P∈R T×N×D ;

[0039] Where T is the number of time steps (e.g., collecting 10 seconds of data at 200 Hz corresponds to T = 2000), N is the number of joints (in this embodiment, N is 14), R is a set of real numbers, and D is the feature of each joint (including joint angle, angular velocity, and acceleration). It can be understood that T also corresponds to the time dimension, N to the joint dimension, and D to the feature dimension.

[0040] Furthermore, the joint motion trajectory matrix and muscle activation timing diagram are input into the trained evaluation model to obtain the coordination efficiency scores and compensatory behavior risk levels of the 14 major joints. In this embodiment, the evaluation model is a GNN (graph neural network) model.

[0041] After obtaining the synergy efficiency scores, compensatory behavior risk levels, and joint mobility levels of the user's 14 major joints, the joint with the lowest synergy efficiency score can be used as the target joint (i.e., the joint to be trained). In an optional embodiment, the user can also select the joint to be trained.

[0042] By evaluating the user's joint mobility and compensatory behavior based on joint angle, angular velocity, and acceleration, the user's joint mobility and compensatory behavior can be clearly identified, thereby ensuring that the subsequent training parameters match the user's actual situation, avoiding the situation where a unified training plan results in mediocre training effects, and thus improving the user's rehabilitation efficiency.

[0043] S3. Determine the training mode based on the risk level of compensatory behavior.

[0044] In this embodiment, the compensatory behavior risk levels include level one risk (low risk), level two risk (medium risk) and level three risk (high risk). The higher the compensatory behavior risk level, the greater the user's tendency to use compensatory behavior.

[0045] Training modes include basic training (i.e., general training) and compensatory behavior-focused training (including weak-link strengthening and forced separation training). The basic training mode provides content related to joint range of motion training, such as performing isometric contractions within the target range of motion (e.g., 60-75°) (e.g., holding the shoulder externally rotated at 65° for 5-8 seconds) to enhance joint stability.

[0046] The weak link strengthening mode allows users to perform resistance training on compensatory areas in addition to joint range of motion training. For example, during lumbar spine training, the patient's pelvis may exhibit compensatory behavior. By limiting pelvic movement and increasing the applied resistance (e.g., by 20%) during training, the compensatory behavior can be reduced and rehabilitation efficiency improved.

[0047] The forced separation mode is: no joint mobility training is provided (that is, the user does not need to reach the target range during training), the abnormal joints are locked, and the target joints are forced to undergo rehabilitation training independently. For example, when performing lumbar spine training, the pelvis is an abnormal part (that is, a compensatory part), then during training, the distance that the pelvis moves is limited (in this embodiment, the movement range is within 2 cm). The training goal of the forced separation mode is not to improve the joint mobility (that is, to perform joint mobility training), but to avoid compensatory behavior as much as possible and to avoid triggering pain to ensure the safety of user training. It should be noted that the forced separation mode needs to be used with the assistance of medical staff.

[0048] In one embodiment, the training mode corresponding to the user's compensatory behavior risk level is determined based on a preset mapping relationship between the compensatory behavior risk level and the training mode. Specifically, if the compensatory behavior risk level is level one, the basic training mode is used; if the compensatory behavior risk level is level two, the weak link reinforcement mode is used; and if the compensatory behavior risk level is level three, the forced separation mode is used.

[0049] By matching the training mode according to the risk level of compensatory behavior of the user's target joint, users with compensatory behavior can try to avoid or limit their compensatory behavior during training. At the same time, the exoskeleton support provides resistance and / or assistance to assist users in effective training, thereby improving training safety and rehabilitation efficiency.

[0050] It should be noted that when some patients cannot meet the training requirements, they need exoskeletons to provide assistance.

[0051] S4. Determine training parameters based on joint mobility risk level.

[0052] In this embodiment, the joint range of motion risk levels include level 1 (low risk), level 2 (medium risk), level 3 (high risk), and level 4 (extremely high risk). Level 1 risk indicates that the functional activities of the target joint are basically unrestricted, and preventive training is mainly used; level 2 risk indicates that daily activities are partially restricted and intensive range of motion (ROM) training is required; level 3 risk indicates severe functional impairment and the risk of injury, requiring medical intervention and daily training; level 4 risk indicates that surgery may be required, and training should be stopped.

[0053] In this embodiment, the training parameters may include resistance, assist, and training speed (a preset standard speed that assists the user in training). Specifically, during the eccentric phase of the exercise, the pneumatic muscles provide resistance (for example, providing resistance when squatting); during the concentric phase of the exercise, the magnetorheological damper provides assist (for example, providing assist when standing). It should be noted that the method for determining assist is similar to that for resistance, and the direction of assist is opposite to that of resistance.

[0054] Specifically, when the joint mobility risk level is a first risk level, the resistance (or assistance) to be provided is determined according to the weight of the user. In this embodiment, the resistance (or assistance) is positively correlated with the weight of the user. Specifically, the weight of the user is multiplied by an adjustment coefficient (set to be less than 1, in this embodiment, 0.3), to obtain the resistance (or assistance) to be provided. It can be understood that the first risk level is generally the case where the joint mobility of the user is close to the minimum mobility in the normal mobility range. For example, the normal range of shoulder external rotation is 60-90°, if the user is exactly 60°, although it is within the normal range, but if no intervention is taken, joint mobility disorder is likely to occur.

[0055] When the joint mobility risk level is a second risk level and a third risk level, the resistance to be provided is determined according to the current joint angle of the target joint. Specifically, the calculation expression of the resistance is:

[0056]

[0057] In the formula, F is the resistance, k is the reference coefficient, T is the time constant, t is the current training duration, ROM is the current joint angle. curent

[0058] Among them, the reference coefficient and the time constant are obtained by deep learning (PPO algorithm), so as to ensure that the resistance can match the actual situation of the user.

[0059] According to the above calculation expression of the resistance, when the user approaches the mobility limit, the ROM current increases but the exponential term starts to saturate, so that the user gets a safe buffer, which can avoid sudden load damage to tissues and excessive traction damage, while ensuring that the user trains within the target intensity, thereby improving the safety and rehabilitation efficiency of the training.

[0060] S5, adjusting the training parameter according to the synergy efficiency score.

[0061] In this embodiment, the reference coefficient is adjusted according to the synergy efficiency score to adjust the resistance to be provided. Specifically, the adjusted reference coefficient is:

[0062]

[0063] In the formula, k is the modified reference coefficient, k_base is the unmodified reference coefficient, and C is the synergy efficiency score.

[0064] ​Furthermore, the training speed is adjusted based on the collaborative efficiency score. Specifically, it is determined whether the collaborative efficiency score is greater than or equal to a score threshold (set to 85 in this embodiment). If so, the current speed is maintained; if not, the collaborative efficiency score is divided by the score threshold, and then multiplied by the current speed to obtain the adjusted speed.

[0065] By correcting the resistance according to the synergistic efficiency score, the user's ineffective compensation during training can be reduced; by correcting the training speed according to the synergistic efficiency score, the problem of user injury caused by speed mismatch can be avoided, thereby improving the safety of user training.

[0066] S6. Output the training mode and the adjusted training parameters.

[0067] Output the training mode and adjusted training parameters to assist users in conducting corresponding training.

[0068] During training, the training parameters are updated every set time (in this embodiment, set to 5 minutes). In one embodiment, the resistance can be updated by updating the reference coefficient. Specifically, the current joint angle is divided by the current reference coefficient, and then multiplied by a preset value (in this embodiment, set to 0.8) to obtain the updated reference coefficient.

[0069] In addition, during training, if the activation intensity of the target muscle group does not reach the preset threshold, the output resistance can be reduced.

[0070] In one embodiment, the method of the present invention further includes: displaying a virtual training scene through AR (augmented reality) technology and outputting prompts.

[0071] Specifically, a three-dimensional skeleton model of the user is constructed using data collected by a depth camera, and the IMU data is aligned with the three-dimensional skeleton model constructed by the depth camera to construct a full-body motion chain.

[0072] Before training, the user's field of view is calibrated through AR glasses, the virtual training scene coordinates are bound, and then the user's ROM threshold is determined based on the virtual scene selected by the user. Specifically, according to the virtual scene selected by the user, it is matched with the preset joint ROM requirement library to obtain the reference ROM threshold (which is the baseline threshold for normal people). For example, if the user selects a rock climbing scene, the shoulder joint external rotation is greater than 70° (ROM threshold). When the user's ROM does not reach 70°, the difficulty is automatically reduced (converted to a simple scene) / threshold.

[0073] During training, AR glasses display a virtual coach and a virtual target line (the virtual target line is determined by the ROM threshold). When the user deviates from the virtual target line, the virtual coach is controlled to output prompts (for example, a voice prompt: "Right knee bends 5 degrees inward, please rotate externally to the green target line").

[0074] In one embodiment, the control method of the present invention further comprises generating a training report, wherein the training report comprises the improvement of joint mobility, the reduction ratio of compensatory movements, and training parameters.

[0075] Figure 2 FIG. 4 is a block diagram schematically illustrating a structure of a whole-body joint rehabilitation training device according to an embodiment of the present invention.

[0076] In a second aspect, the present invention also provides a whole body joint rehabilitation training device. Figure 2 As shown in the figure, the whole-body joint rehabilitation training device includes flexible sensors, an IMU array, high-density surface electromyography (EMG), an exoskeleton support, and a controller. Specifically, the flexible sensors are embedded in elastic fabric, covering the 14 major joints of the human body, and are used to collect joint angles; the IMU array uses an MPU-9250 module (integrated accelerometer + gyroscope), which is deployed at the protruding points of bones near the joints to collect acceleration and angular velocity in real time; the high-density surface electromyography (HD-sEMG) has 16-channel electrode patches, which are attached to the target muscle groups (such as the quadriceps and deltoid muscles) to collect the EMG signals of the target muscle groups.

[0077] Exoskeleton (can be Figure 3 The exoskeleton support in the prior art shown includes pneumatic artificial muscles (PAM) and magnetorheological dampers to provide the resistance and power required for training;

[0078] The controller is connected to the flexible sensor, the IMU array, the high-density surface electromyography and the exoskeleton bracket, and is used to implement the control method of the whole-body joint rehabilitation training device described in the first aspect.

[0079] In other optional embodiments, AR glasses may also be included.

[0080] During use, the user wears flexible sensor clothing that fits the joints and muscle groups. The exoskeleton bracket adjusts its length according to the user's body shape. The AR glasses calibrate the user's field of view and bind the virtual training scene coordinate system. The controller adjusts the training parameters according to the user's sensor data.

[0081] In the description of this specification, "multiple" means at least two, such as two, three or more, etc., unless otherwise specifically defined. In addition, the steps of the above method are divided only for the purpose of clarity of description. When implementing, they can be combined into one step or some steps can be split and decomposed into multiple steps, as long as they include the same logical relationship.

[0082] While several embodiments of the present invention have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Numerous modifications, variations, and alternatives will occur to those skilled in the art without departing from the concept and spirit of the present invention. It should be understood that various alternatives to the embodiments of the present invention described herein may be employed in practicing the present invention.

Claims

1. A control method for a whole-body joint rehabilitation training device, characterized in that: include: Collect joint angles, IMU data, and electromyographic signals when the user performs multiple standard movements; determining a joint mobility risk level of the target joint based on the joint angle; determining a compensatory behavior risk level and a synergy efficiency score of a target joint based on the joint angle, the IMU data, and the electromyographic signal, wherein the compensatory behavior risk level represents the user's tendency to utilize compensatory behavior, and the synergy efficiency score represents the synergy between the target joint and other joints; Determining a training mode according to the compensatory behavior risk level, wherein the training mode includes at least a general mode and a mode targeting compensatory behavior; determining training parameters according to the joint mobility risk level, wherein the training parameters include resistance; adjusting the training parameters according to the synergy efficiency score, wherein the synergy efficiency score is positively correlated with the training parameters; The training mode and the adjusted training parameters are output.

2. The control method of the whole body joint rehabilitation training device according to claim 1, characterized in that: Determining a training mode according to the compensatory behavior risk level includes: determining the training mode according to a preset mapping relationship between the compensatory behavior risk level and the training mode.

3. The control method of the whole body joint rehabilitation training device according to claim 1, characterized in that: Determining training parameters according to the joint mobility risk level includes: judging whether the joint mobility risk level is a level one risk; if so, determining the resistance to be provided according to the user's weight, and the resistance is positively correlated with the weight; if not, determining the resistance to be provided according to the current joint angle of the target joint, and the resistance is positively correlated with the current joint angle.

4. The control method of the whole body joint rehabilitation training device according to claim 3, characterized in that: Determine the resistance required based on the current range of motion, including: Where F is the resistance to be provided, k is the reference coefficient, T is the time constant, t is the current training time, ROM curent is the current joint angle.

5. The control method of the whole body joint rehabilitation training device according to claim 4, characterized in that: Adjusting the training parameters according to the synergistic efficiency score includes: adjusting the benchmark coefficient according to the synergistic efficiency score to adjust the resistance, and the benchmark coefficient is positively correlated with the synergistic efficiency score.

6. The control method of the whole body joint rehabilitation training device according to claim 1, characterized in that: The training process also includes: adjusting the training parameters at set time intervals.

7. The control method of the whole body joint rehabilitation training device according to claim 1, characterized in that: Determining a compensatory behavior risk level and a synergistic efficiency score of a target joint based on the joint angle, the IMU data, and the electromyographic signal, including: The joint angle and the IMU data are fused to obtain a joint motion trajectory matrix; and a muscle activation timing diagram is obtained based on the electromyographic signal; The joint motion trajectory matrix and the muscle activation timing diagram are input into a trained evaluation model to obtain the synergy efficiency score and the compensatory behavior risk level; the evaluation model is a machine learning model.

8. The control method of the whole body joint rehabilitation training device according to claim 1, characterized in that: Determine the joint range of motion risk level of the target joint based on the joint angle, including: matching a standard range of motion according to the target joint; The joint mobility risk level is determined based on the difference between the joint angle range of the user's target joint and the standard range of motion.

9. The control method of the whole body joint rehabilitation training device according to claim 1, characterized in that: Also includes: The virtual training scene is displayed through AR technology, and corresponding prompts are output.

10. A whole body joint rehabilitation training device, characterized in that: include Flexible sensors, IMU arrays, and high-density surface electromyography (SEM), wherein the flexible sensors are used to collect joint angles, the IMU arrays are used to collect IMU data, and the high-density surface electromyography (SEM) is used to collect electromyographic signals; An exoskeleton support, wherein the exoskeleton support is used to provide resistance and assistance required for training; A controller is connected to the flexible sensor, the IMU array, the high-density surface electromyography, and the exoskeleton bracket, and is used to implement the control method of the whole-body joint rehabilitation training device described in any one of claims 1-9.

Images