A bilateral upper limb rehabilitation training system and method based on goal orientation

Abstract

The present invention discloses a bilateral upper limb rehabilitation training system and method based on goal-oriented. The system includes: a posture acquisition module, which is used to collect the actual posture information of the patient's healthy arm; a host computer module, which is used to give the target posture information and the assist torque level of the patient's affected arm; a rehabilitation robot module, which is used to control the patient's affected arm to perform rehabilitation exercises, obtain the actual posture information of the patient's affected arm and the actual posture information of the rehabilitation robot; a lower computer module, which is used to receive and calculate the target human-computer interaction torque and actual human-computer interaction torque of the patient's affected arm according to the actual posture information, target posture information, assist torque level and actual posture information of the patient's healthy arm, and introduce a bilateral mechanism. The system can effectively utilize the healthy side-related data of the patient during training, reduce the fatigue of the patient's healthy side caused by training, and improve the patient's enthusiasm for active training.

Description

Technical Field

[0001] The present invention belongs to the field of rehabilitation training, and in particular relates to a bilateral upper limb rehabilitation training system and method based on goal orientation. Background Art

[0002] Stroke is a common acute cerebrovascular disease that often causes damage to the nerve center and has the characteristics of high morbidity and high disability rate. The rehabilitation training system installed on the rehabilitation robot can effectively assist stroke patients in rehabilitation training, thereby promoting the remodeling of damaged nerves.

[0003] Typical bilateral collaborative upper limb rehabilitation robots in the prior art usually adopt a mirror training mode, allowing the patient's healthy upper limb to participate in the rehabilitation training. On the one hand, through the combination of virtual reality, the patient can observe the movement of his healthy side and imagine the movement of his affected side, which can enhance the patient's intention to actively exercise the affected limb, thereby contributing to the rehabilitation of the patient's motor function on the affected side. On the other hand, by analyzing the posture, joint angle and other data of the patient's healthy upper limb as the basis for the control system to calculate the human-computer interaction torque of the affected arm, the exoskeleton can be driven to provide assistance for the patient's affected upper limb to complete rehabilitation training.

[0004] For example, CN112206124A discloses a neural circuit-guided upper limb function rehabilitation training system and method, which specifically describes a virtual mirror training mode, in which a virtual contralateral limb movement image is displayed in a 3D virtual reality scene of a display module, and a driving device is controlled by an analysis and control module to help patients train; another example is CN113397918A discloses a wearable elbow joint exoskeleton rehabilitation control system, which specifically describes a mirror training mode, by collecting the angle signal of the patient's healthy upper limb, analyzing the elbow joint angle of the healthy upper limb, and controlling the motor module to drive the exoskeleton according to the elbow joint angle to drive the patient's diseased upper limb to perform synchronous flexion and extension rehabilitation movements following the healthy upper limb.

[0005] The above existing technical solutions have the following disadvantages: when the patient performs bilateral coordinated training, the healthy upper limb must always perform synchronous movement. This bilateral real-time synchronous movement is likely to increase the fatigue of the healthy side of the patient during rehabilitation training.

[0006] On the other hand, in the existing technical solutions, the control system calculates the human-computer interaction torque required by the patient's affected side in real time based on the patient's healthy upper limb posture information to provide assistance for the affected side training. However, the patient does not need help in real time when performing rehabilitation training on the affected side. Especially for patients who need active training, real-time assistance monitoring and compensation will increase the patient's dependence on the drive device and reduce the patient's enthusiasm for active training.

[0007] Therefore, in bilateral coordinated upper limb rehabilitation training, how to effectively utilize the relevant data of the healthy side during the patient's training process to reduce the fatigue of the patient's healthy side caused by training while increasing the patient's enthusiasm for active training is an urgent problem to be solved by technical personnel in this field. Summary of the invention

[0008] In order to solve at least one of the problems existing in the prior art, the present invention provides a goal-oriented bilateral upper limb rehabilitation training system, which can effectively utilize the posture information of the healthy side of the patient during the training process, introduce the movement of the patient's healthy side only when the patient's affected side has difficulties in the process of goal-oriented active training, and use a driving device to provide assistance to the affected side.

[0009] To achieve the purpose of the present invention, the present invention provides a goal-oriented bilateral upper limb rehabilitation training system comprising:

[0010] The posture acquisition module is used to collect the actual posture information of the patient's healthy arm;

[0011] The host computer module is used to provide the target posture information and assist torque level of the patient's affected arm;

[0012] The rehabilitation robot module is used to control the patient's affected arm to perform rehabilitation exercises and obtain the actual posture information of the patient's affected arm and the actual posture information of the rehabilitation robot;

[0013] The lower computer module is used to receive the actual posture information of the patient's affected arm, the target posture information of the patient's affected arm, the assist torque level and the actual posture information of the patient's healthy arm, and calculate the target human-computer interaction torque and the actual human-computer interaction torque of the patient's affected arm; and

[0014] The lower computer module introduces a bilateral mechanism in the traditional goal-oriented active training to calculate the target human-computer interaction torque of the patient's affected arm. The actual position information of the patient's affected arm Target posture information closer to the patient's affected arm When providing torque level T based on the boost ass The assistance to the target posture information of the patient's affected arm is the target human-machine interaction torque T of the patient's affected arm. d , otherwise no assistance is provided, and the formula of the bilateral mechanism is as follows:

[0015]

[0016] Among them, T d is the target human-machine interaction torque of the patient's affected arm, is the target posture information of the patient's affected arm, is the actual position information of the patient's affected arm, is the actual position information of the patient’s healthy arm, T ass The function Sign is a sign function, which returns 1 if it is positive and -1 if it is negative.

[0017] Preferably, the posture acquisition module includes any one of a posture sensor, a visual sensor, and a joint angle sensor.

[0018] Preferably, the host computer module is further provided with:

[0019] A user interface module, used for receiving the actual posture of the patient's affected arm, and generating and displaying the target posture information and assist torque level of the patient's affected arm in real time;

[0020] The game training module is used to regulate the patient's rehabilitation training tasks and enhance the interest of the patient's training.

[0021] Preferably, the game training module is equipped with a rehabilitation training game. The upper computer module receives the actual posture information of the patient's affected arm sent by the lower computer module as input to the rehabilitation training game. The target posture information of the patient's affected arm is randomly generated in real time by the rehabilitation training game with reference to standard rehabilitation training movements.

[0022] Preferably, the lower computer module calculates the actual human-computer interaction torque T of the patient's affected arm by the following formula: h , the specific formula is shown in the following formulas (2) to (5):

[0023] T h =T c -T g (2)

[0024]

[0025] In the above formula, T g The actual torque of each joint caused by the weight of the rehabilitation robot module, T c is the actual torque of each joint of the rehabilitation robot module, m p is the mass of each joint in the rehabilitation robot module, g T is the gravitational acceleration vector, is the position vector of each joint’s center of mass relative to the joint coordinate system, is the actual posture information of the rehabilitation robot, p is the maximum number of joint links, A 0r , A 1r , ..., A ir , A pr are the homogeneous transformation matrices of each joint link of the rehabilitation robot, such as A iris the homogeneous transformation matrix of the i-th joint link, and so on.

[0026] Preferably, the rehabilitation robot module implements control of the rehabilitation training of the patient's affected arm through an admittance control algorithm.

[0027] Preferably, the specific formula of the admittance control algorithm is as shown in the following formula (6):

[0028]

[0029] Among them, in the above formula is the control signal of the input motor, M d is the target inertia matrix, D d is the target damping matrix, T h is the actual human-machine interaction torque of the patient's affected arm.

[0030] Preferably, the target inertia matrix M in the formula of the admittance control algorithm is d Set it to 0 matrix, and get:

[0031]

[0032] Then we have:

[0033]

[0034] The present invention also provides a goal-oriented bilateral upper limb rehabilitation training method, which is implemented based on the above-mentioned system and includes the following steps:

[0035] Collecting actual posture information of the healthy arm of the patient, collecting actual posture information of the rehabilitation robot module and actual posture information of the affected arm of the patient, giving target posture information of the affected arm of the patient and giving the assist torque level of the affected arm of the patient;

[0036] The target human-machine interaction torque of the patient's affected arm is obtained by analyzing the actual posture information of the patient's affected arm, the target posture information of the patient's affected arm, the assist torque level and the actual posture information of the patient's healthy arm;

[0037] The rehabilitation robot module receives the target human-machine interaction torque and pulls the patient's affected arm to perform rehabilitation exercises. When the affected side is actively trained, the actual posture information of the patient's healthy arm is The actual position information of the patient's affected arm Target posture information closer to the patient's affected arm When providing torque level T based on the boost ass The assistance to the target posture information of the patient's affected arm is the target human-machine interaction torque T of the patient's affected arm. d, otherwise no assistance will be provided.

[0038] Compared with the prior art, the present invention has at least the following advantages:

[0039] The present invention proposes a goal-oriented bilateral upper limb rehabilitation training system, which effectively utilizes the posture information of the healthy side of the patient during the training process, introduces the movement of the patient's healthy side only when the patient's affected side has difficulties in the process of goal-oriented active training, and uses a driving device to provide assistance to the affected side to reduce the fatigue of the patient's healthy side caused by training, while improving the patient's enthusiasm for active training. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A schematic diagram of the structure of a goal-oriented bilateral upper limb rehabilitation training system provided by an embodiment of the present invention;

[0041] Figure 2 It is a schematic diagram of the steps of the goal-oriented bilateral upper limb rehabilitation training method according to an embodiment of the present invention;

[0042] Figure 3 It is a comparison diagram of the real-time trajectory of the shoulder flexion and extension angle and the actual assist torque in active and passive training according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0044] The embodiment of the present invention provides a goal-oriented bilateral upper limb rehabilitation training system and method, which can effectively utilize the posture information of the healthy side of the patient during the training process, and introduce the movement of the patient's healthy side only when the patient's affected side has difficulties in the process of goal-oriented active training, and use the driving device to provide assistance to the affected side.

[0045] See also Figure 1 , Figure 1 A schematic structural diagram of a goal-oriented bilateral upper limb rehabilitation training system provided in an embodiment of the present invention.

[0046] The present invention provides a goal-oriented bilateral upper limb rehabilitation training system, comprising a posture acquisition module 1, a host computer module 2, a lower computer module 3 and a rehabilitation robot module 4, wherein the posture acquisition module 1 is used to acquire actual posture information of a patient's healthy arm, and the host computer module 2 is used to give target posture information and an assist torque level of the patient's affected arm; the rehabilitation robot module 4 is used to be worn on the patient's upper limb to control the patient's affected arm to perform rehabilitation exercises, and obtain the actual posture information of the patient's affected arm and the actual posture information of the rehabilitation robot; the lower computer module 3 is used to receive and calculate the target human-computer interaction torque and actual human-computer interaction torque of the patient's affected arm based on the actual posture information of the patient's affected arm, the target posture information of the patient's affected arm, the assist torque level and the actual posture information of the patient's healthy arm.

[0047] In some embodiments of the present invention, the posture acquisition module 1 includes three IMU posture sensors, which are respectively attached to the upper arm, forearm and trunk of the patient's healthy upper limb with straps. The actual posture information of the patient's healthy arm is obtained by analyzing the Euler angle information obtained by the three IMU posture sensors.

[0048] The rehabilitation robot module 4 adopts an existing rehabilitation robot. In some embodiments of the present invention, for example, an upper limb rehabilitation exoskeleton disclosed in Chinese invention patent CN116059079A can be adopted. The rehabilitation robot module 4 is a four-degree-of-freedom upper limb rehabilitation exoskeleton system, including a shoulder (three degrees of freedom) and an elbow (one degree of freedom). The shoulder and elbow joint motors of the upper limb rehabilitation exoskeleton are all equipped with built-in angle encoders, which can provide real-time feedback on the actual posture information of the upper limb rehabilitation exoskeleton. The upper limb rehabilitation exoskeleton is tied to the patient's affected arm, and is used to drive the patient's affected arm in real time and detect the actual posture information of the patient's affected arm. The shoulder and elbow joints of the upper limb rehabilitation exoskeleton are equipped with built-in torque sensors, which can detect and provide real-time feedback on the actual human-machine interaction torque information of the patient's affected arm.

[0049] In some embodiments of the present invention, the host computer module includes:

[0050] A user interface module, used for receiving the actual posture information of the patient's affected arm, and generating and displaying the target posture information of the patient's affected arm and the assist torque level in real time;

[0051] The game training module is used to regulate the patient's rehabilitation training tasks and enhance the interest of the patient's training.

[0052] The host computer module 2 is equipped with a rehabilitation training game, which provides the target posture information and assist torque level of the patient's affected arm.

[0053] In some embodiments of the present invention, the lower computer module 3 is used to control the movement of the rehabilitation robot module 4. The lower computer module 3 receives the actual posture information of the patient's healthy arm, the actual posture information of the patient's affected arm, the target posture information of the patient's affected arm and the assist torque level, and obtains the target human-computer interaction torque of the patient's affected arm through analysis and calculation. The lower computer module 3 sends the actual posture information of the patient's healthy arm and the actual posture information of the patient's affected arm to the upper computer module 2 as input for the rehabilitation training game.

[0054] See also Figure 2 , is a schematic diagram of a workflow of a goal-oriented bilateral upper limb rehabilitation training method provided by an embodiment of the present invention, wherein the method comprises the following steps:

[0055] S101: Collecting the actual posture information of the healthy arm of the patient, collecting the actual posture information of the rehabilitation robot module 4 and the actual posture information of the affected arm of the patient, giving the target posture information of the affected arm of the patient and giving the assist torque level of the affected arm of the patient.

[0056] In some embodiments of the present invention, the actual posture information of the healthy arm of the patient is obtained by a posture acquisition module 1, which includes three IMU posture sensors, which are respectively attached to the upper arm, forearm and body trunk of the healthy upper limb of the patient with straps. The actual posture information of the healthy arm of the patient is obtained by analyzing the Euler angle information obtained by the three IMU posture sensor modules 1, which is recorded as The actual posture information of the patient's healthy arm is output to the CAN bus via CAN communication.

[0057] In some embodiments of the present invention, the actual posture information of the rehabilitation robot module 4 and the actual posture information of the affected arm of the patient are obtained through an upper limb rehabilitation exoskeleton. Preferably, as disclosed in CN116059079A, in an upper limb rehabilitation exoskeleton system, a total of 4 motors in the left and right shoulders and elbows control the movement of their respective plates through independent drive plates. The control mode of each motor includes a position control mode, a speed control mode and a torque control mode. Among them, the position control mode is used to accurately locate the position of the motor, the speed control mode is used to adjust the speed of the motor, and the torque control mode is used to control the torque output by the motor; the output ends of the four motors are equipped with torque sensors, which can monitor the torque output by the corresponding motors in real time, and feed back the data to the main control system of the upper limb rehabilitation exoskeleton for adjusting and optimizing the operating parameters of the motors to ensure that the movement of the upper limb rehabilitation exoskeleton meets the requirements of rehabilitation training. Preferably, the data of these torque sensors are the actual torque of each joint of the upper limb rehabilitation exoskeleton, denoted as T cThe upper limb rehabilitation exoskeleton is strapped to the patient's affected arm through a strap, and the actual posture information of the upper limb rehabilitation exoskeleton (denoted as ) and the actual posture information of the patient's affected arm (denoted as ), and output the information to the CAN bus via CAN communication.

[0058] In some embodiments of the present invention, the target posture information of the patient's affected arm is obtained in real time by the rehabilitation training game carried by the host computer module 2. Preferably, the target posture information of the patient's affected arm is randomly generated in real time by the rehabilitation training game with reference to standard rehabilitation training movements.

[0059] In some embodiments of the present invention, the assist torque level of the patient's affected arm is preset by the therapist and sent to the lower computer module 3 through the upper computer module 2 to adjust the assist torque level. The upper computer module 2 receives the actual posture information of the patient's affected arm sent by the lower computer module 3 as input for the rehabilitation training game. The upper computer module 2 sends the target posture information and assist torque level of the patient's affected arm to the lower computer module 3.

[0060] S102: Target human-machine interaction torque of the patient's affected arm is obtained by analyzing actual posture information of the patient's affected arm, target posture information of the patient's affected arm, assist torque level and actual posture information of the patient's healthy arm.

[0061] In some embodiments of the present invention, in order to analyze the actual posture information of the patient's affected arm, the target posture information of the affected arm, the assist torque level and the actual posture information of the patient's healthy arm, and obtain the target human-machine interaction torque of the patient's affected arm, a bilateral mechanism is introduced in the goal-oriented active training, and the bilateral mechanism is shown in formula (1):

[0062]

[0063] Among them, T d is the target human-machine interaction torque of the patient's affected arm, is the target posture information of the patient's affected arm, is the actual position information of the patient's affected arm, is the actual position information of the patient’s healthy arm, T ass The function Sign is a sign function, which returns 1 if it is positive and -1 if it is negative.

[0064] The specific principles are:

[0065] When the actual position information of the patient's healthy arm The actual position information of the patient's affected arm Target posture information closer to the patient's affected arm When providing torque level T based on the boost ass The assistance to the target posture information of the patient's affected arm is the target human-machine interaction torque T of the patient's affected arm. d , otherwise no assistance is provided. In a specific rehabilitation training scenario, it can be understood that when a patient is undergoing active rehabilitation training and the movement of his affected arm cannot reach the target posture, the patient's healthy arm joins the training and the healthy arm performs training synchronously with the patient's affected arm. When the actual posture information of the patient's healthy arm The actual position information of the patient's affected arm Target posture information closer to the patient's affected arm When providing torque level T based on the boost ass The assistance to the target posture information of the patient's affected arm is the target human-machine interaction torque T of the patient's affected arm. d , so that the patient's affected arm can reach the target and complete the training.

[0066] In some embodiments of the present invention, the actual human-machine interaction torque T h Calculated from formula (2).

[0067] T h =T c -T g (2)

[0068] Where, T g The actual torque of each joint caused by the weight of the rehabilitation robot module, T c is the actual torque of each joint of the rehabilitation robot module.

[0069] In order to accurately calculate the actual torque T of each joint caused by the weight of the upper limb rehabilitation exoskeleton (rehabilitation robot module), g , the torque equations of each joint of the robot are obtained by using the Lagrangian function-based method, as shown in equations (3) to (5). In this process, considering that the movement speed of the upper limb rehabilitation exoskeleton is relatively slow, the influence of the inertia term, Coriolis force and centripetal force term are ignored in the system dynamics modeling.

[0070]

[0071] In the formula, m p is the mass of each joint in the upper limb rehabilitation exoskeleton (rehabilitation robot module), g T is the gravitational acceleration vector, is the actual posture information of the upper limb rehabilitation exoskeleton, p is the maximum number of joint links of the rehabilitation robot module, A 0r , A1r , ..., A ir , A pr are the homogeneous transformation matrices of each joint link of the rehabilitation robot module, such as A ir is the homogeneous transformation matrix of the ith joint link, and so on. In the embodiment of the present invention, there are 4 joints and 4 links, is the position vector of the center of mass of each joint relative to the joint coordinate system, and the joint coordinate system is expressed as the homogeneous transformation matrix of the corresponding joint.

[0072] S103: The rehabilitation robot module receives the target human-machine interaction torque and pulls the patient's affected arm to perform rehabilitation exercises.

[0073] In this step, after the upper limb rehabilitation exoskeleton receives the target human-computer interaction torque of the affected arm, it controls the motor of the upper limb rehabilitation exoskeleton to operate so that the upper limb rehabilitation exoskeleton can pull the patient's affected arm to complete bilateral active and passive rehabilitation training.

[0074] In some embodiments of the present invention, the control of rehabilitation training for the affected arm of the patient is achieved by an admittance control algorithm, and the specific formula of the admittance control algorithm is shown in the following formula (6):

[0075]

[0076] In the formula, are the position, speed and acceleration control signals of the input motor respectively, M d is the target inertia matrix, D d is the target damping matrix, T h is the actual human-machine interaction torque of the patient's affected arm.

[0077] Preferably, since the upper limb rehabilitation exoskeleton moves slowly, in order to simplify the control model, the target inertia matrix M d Set it to 0 matrix, and we can get formula (7). Specifically, after further derivation, we can get the control rate of each motor of the upper limb rehabilitation exoskeleton, as shown in formula (8):

[0078]

[0079]

[0080] In some embodiments of the present invention, a goal-oriented bilateral upper limb rehabilitation training method is provided, which is implemented based on the system provided in the above embodiments, and the method includes the following steps:

[0081] Collecting actual posture information of the healthy arm of the patient, collecting actual posture information of the rehabilitation robot module and actual posture information of the affected arm of the patient, giving target posture information of the affected arm of the patient and giving the assist torque level of the affected arm of the patient;

[0082] The target human-machine interaction torque of the patient's affected arm is obtained by analyzing the actual posture information of the patient's affected arm, the target posture information of the patient's affected arm, the assist torque level and the actual posture information of the patient's healthy arm;

[0083] The rehabilitation robot module receives the target human-machine interaction torque and pulls the patient's affected arm to perform rehabilitation exercises. When the affected side is actively trained, the actual posture information of the patient's healthy arm is The actual position information of the patient's affected arm Target posture information closer to the patient's affected arm When providing torque level T based on the boost ass The assistance to the target posture information of the patient's affected arm is the target human-machine interaction torque T of the patient's affected arm. d , otherwise no assistance will be provided.

[0084] In some embodiments of the present invention, the comparison diagram of the shoulder flexion and extension angle and the actual assist torque real-time trajectory during active and passive training is as follows: Figure 3 As shown. Figure 3 It can be seen that from 22.9s to 29.4s, the actual shoulder flexion / extension angle of the affected side is approaching the target shoulder flexion / extension angle of the affected side, that is, the subject is trying to complete the goal-oriented active training. At 26.5s, the subject simulated the situation that the patient was insufficiently strong and the affected side could not continue to lift up. At this time, the subject's healthy side was slowly lifted until 29.4s, when the actual shoulder flexion / extension angle of the healthy side exceeded the actual shoulder flexion / extension angle of the affected side. The upper limb rehabilitation exoskeleton system provided an actual assist torque of 6N·m. The corresponding actual assist torque trajectory is shown as follows: Figure 3 As shown in (c). At this time, since the upper limb rehabilitation exoskeleton system provides assistance to the affected side, the affected side can continue to lift up and eventually reach the target shoulder flexion / extension angle of the affected side. In the rehabilitation scenario of goal-oriented active training, once the actual healthy side shoulder flexion / extension angle reaches the target healthy side shoulder flexion / extension angle, the target healthy side shoulder flexion / extension angle will be updated, and the actual assist torque will be updated to 0 N·m. Therefore, the goal-oriented bilateral upper limb rehabilitation training system and method provided by the aforementioned embodiments of the present invention can effectively utilize the healthy side posture information of the patient during training to provide assistance to the patient.

[0085] The above is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as above in the preferred embodiment, it is not used to limit the present invention. Any technician familiar with the art can make many possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above without departing from the scope of the technical solution of the present invention, or modify it into an equivalent embodiment of equivalent changes. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of protection of the technical solution of the present invention.

Claims

1. A goal-oriented bilateral upper limb rehabilitation training system, characterized in that: include: The posture acquisition module is used to collect the actual posture information of the patient's healthy arm; The host computer module is used to provide the target posture information and assist torque level of the patient's affected arm; The rehabilitation robot module is used to control the patient's affected arm to perform rehabilitation exercises and obtain the actual posture information of the patient's affected arm and the actual posture information of the rehabilitation robot; The lower computer module is used to receive the actual posture information of the patient's affected arm, the target posture information of the patient's affected arm, the assist torque level and the actual posture information of the patient's healthy arm, and calculate the target human-computer interaction torque and the actual human-computer interaction torque of the patient's affected arm; and The lower computer module introduces a bilateral mechanism in the traditional goal-oriented active training to calculate the target human-computer interaction torque of the patient's affected arm. The actual position information of the patient's affected arm Target posture information closer to the patient's affected arm When providing torque level T based on the boost ass The assistance to the target posture information of the patient's affected arm is the target human-machine interaction torque T of the patient's affected arm. d , otherwise no assistance is provided, and the formula of the bilateral mechanism is as follows: Among them, T d is the target human-machine interaction torque of the patient's affected arm, is the target posture information of the patient's affected arm, is the actual position information of the patient's affected arm, is the actual position information of the patient’s healthy arm, T ass is the assist torque level, and the function Sign is the sign function.

2. A goal-oriented bilateral upper limb rehabilitation training system according to claim 1, characterized in that: The posture acquisition module includes any one of a posture sensor, a visual sensor, and a joint angle sensor.

3. A goal-oriented bilateral upper limb rehabilitation training system according to claim 1, characterized in that: The host computer module includes: A user interface module, used for receiving the actual posture of the patient's affected arm, and generating and displaying the target posture information of the patient's affected arm and the assist torque level in real time; Game training module, used to regulate the patient's rehabilitation training tasks.

4. A goal-oriented bilateral upper limb rehabilitation training system according to claim 3, characterized in that: The game training module is equipped with a rehabilitation training game. The upper computer module receives the actual posture information of the patient's affected arm sent by the lower computer module as the input of the rehabilitation training game. The target posture information of the patient's affected arm is randomly generated in real time by the rehabilitation training game with reference to standard rehabilitation training movements.

5. The goal-oriented bilateral upper limb rehabilitation training system according to claim 1, characterized in that: The lower computer module calculates the actual human-computer interaction torque T of the patient's affected arm by the following formula h , the formula is as follows: T h =T c -T g Among them, T g is the actual torque of each joint caused by the weight of the rehabilitation robot module, T c is the actual torque of each joint of the rehabilitation robot module.

6. A goal-oriented bilateral upper limb rehabilitation training system according to claim 5, characterized in that: The actual joint torque T caused by the weight of the rehabilitation robot module g The calculation method is: Among them, m p is the mass of each joint in the rehabilitation robot module, g T is the gravitational acceleration vector, is the position vector of each joint’s center of mass relative to the joint coordinate system, is the actual posture information of the rehabilitation robot module, p is the maximum number of joint links, A 0r , A 1r , ..., A ir , A pr are the homogeneous transformation matrices of each joint link of the rehabilitation robot module, A ir is the homogeneous transformation matrix of the i-th joint link.

7. A goal-oriented bilateral upper limb rehabilitation training system according to any one of claims 1 to 6, characterized in that: The rehabilitation robot module implements control of the rehabilitation training of the patient's affected arm through an admittance control algorithm.

8. The goal-oriented bilateral upper limb rehabilitation training system according to claim 6, characterized in that: The formula of the admittance control algorithm is as follows: in, are the position, speed and acceleration control signals of the input motor respectively, M d is the target inertia matrix, D d is the target damping matrix.

9. A goal-oriented bilateral upper limb rehabilitation training system according to claim 8, characterized in that: The target inertia matrix M in the formula of the admittance control algorithm is d Set it to 0 matrix, and get: Then we have:

10. A goal-oriented bilateral upper limb rehabilitation training method, characterized in that: Based on the system implementation of claims 1-9, the method comprises the following steps: Collecting actual posture information of the healthy arm of the patient, collecting actual posture information of the rehabilitation robot module and actual posture information of the affected arm of the patient, giving target posture information of the affected arm of the patient and giving the assist torque level of the affected arm of the patient; The target human-machine interaction torque of the patient's affected arm is obtained by analyzing the actual posture information of the patient's affected arm, the target posture information of the patient's affected arm, the assist torque level and the actual posture information of the patient's healthy arm; The rehabilitation robot module receives the target human-machine interaction torque and pulls the patient's affected arm to perform rehabilitation exercises. When the affected side is actively trained, the actual posture information of the patient's healthy arm is The actual position information of the patient's affected arm Target posture information closer to the patient's affected arm When providing torque level T based on the boost ass The assistance to the target posture information of the patient's affected arm is the target human-machine interaction torque T of the patient's affected arm. d , otherwise no assistance will be provided.

Images