Trajectory motion planning method for rehabilitation training robots and rehabilitation training robots

Abstract

This invention relates to the field of rehabilitation robot technology, specifically to a trajectory motion planning method for rehabilitation training robots and a rehabilitation training robot itself. It solves the technical problem that existing rehabilitation training robots struggle to adapt their movements to various patient types, leading to high difficulty in rehabilitation training and a high risk of secondary injury. The invention provides a trajectory motion planning method for a rehabilitation training robot, constructing a trajectory template library that includes trajectory templates for different motor function levels and training positions. When performing rehabilitation training on a user, the reference motion trajectory used by the robotic arm is matched according to the user's current motor function level. Therefore, using a matched reference motion trajectory for rehabilitation training reduces the difficulty of rehabilitation training and decreases the risk of secondary injury.

Description

Technical Field

[0001] This invention relates to the field of rehabilitation robot technology, and more specifically, to a trajectory motion planning method for rehabilitation training robots and a rehabilitation training robot. Background Technology

[0002] Many patients with stroke, traumatic brain injury, and spinal cord injury suffer from upper limb motor dysfunction, severely impacting their quality of life. These patients require active rehabilitation therapy to restore their function. Rehabilitation robots can assist or even replace physicians in providing patients with more continuous, effective, and targeted rehabilitation training and treatment. Furthermore, they can record patient treatment data in real time, providing objective evidence for condition assessment and treatment plan improvement.

[0003] Current rehabilitation robots typically learn their training trajectories from healthy individuals through a teaching method. The operator directly grasps the robot's end effector (such as a handle) or linkage, using their professional knowledge and experience to manually guide the robot to simulate the desired rehabilitation trajectory template for the patient. Throughout this process, the robot's built-in sensors record the position, velocity, and even force information of all its joints at a high frequency. This data is processed to form a digital trajectory template that the robot can repeatedly execute. When using the robot for patient rehabilitation training, the robot's trajectory template is based on a pre-stored template corresponding to a healthy individual. However, for individuals with motor function impairments, upper limb movement has unique pathological characteristics, and the trajectory of a healthy person may not be suitable. Especially for those with poor motor function, directly using the trajectory of a healthy person for rehabilitation training is too difficult and carries the risk of secondary injury. Summary of the Invention

[0004] The purpose of this invention is to provide a trajectory motion planning method for a rehabilitation training robot and a rehabilitation training robot, which solves the technical problem that the rehabilitation movements of existing rehabilitation training robots are difficult to adapt to various types of patients, resulting in high difficulty in rehabilitation training and high risk of secondary injury.

[0005] As a first aspect of the present invention, the present invention provides a trajectory motion planning method for a rehabilitation training robot, comprising:

[0006] A trajectory template library for the robotic arm of a rehabilitation training robot is constructed. The trajectory template library stores trajectory templates and the corresponding motor function levels of the trajectory templates. The trajectory templates represent the trajectory of a person of the corresponding motor function level moving from an initial position in the movement space to a training position in the movement space.

[0007] Based on the target training location selected by the user, a reference motion trajectory matching the user's current motor function level is searched in the trajectory template library;

[0008] The robotic arm is controlled to move to the target training position in the motion space according to the reference motion trajectory in order to perform rehabilitation training on the user.

[0009] In one embodiment of the present invention, the robotic arm includes multiple degrees of freedom for the robotic arm's moving parts, and the multiple degrees of freedom for the robotic arm's moving parts correspond one-to-one with the degrees of freedom of the human upper limb joints, including the shoulder joint's external swing or adduction degree of freedom, shoulder joint's flexion or extension degree of freedom, shoulder joint's internal rotation or external rotation degree of freedom, elbow joint's flexion or extension degree of freedom, forearm's pronation or supination degree of freedom, wrist joint's ulnar flexion or radial flexion degree of freedom, and wrist joint's dorsiflexion or palmar flexion degree of freedom.

[0010] The trajectory template includes multiple sub-trajectory templates. The sub-trajectory templates are time series of upper limb joint degree-of-freedom parameters during the process of a person of a corresponding motor function level moving from an initial position in the movement space to a training position in the movement space. The joint degree-of-freedom parameters include joint angles and angular velocities.

[0011] In one embodiment of the present invention, controlling the robotic arm to move to the target training position in the motion space according to the reference motion trajectory includes:

[0012] Analyze multiple sub-reference trajectories of the reference motion trajectory;

[0013] The movement of the robotic arm's moving parts corresponding to the sub-reference trajectory is controlled according to the sub-reference trajectory.

[0014] In one embodiment of the present invention, before searching for a reference motion trajectory in the trajectory template library that matches the user's current motor function level based on the target training position selected by the user, the trajectory motion planning method further includes:

[0015] Acquire the cognitive tasks included in the rehabilitation training, and display multiple preset information corresponding to the cognitive tasks at corresponding preset positions on the display device;

[0016] The eye tracker is controlled to capture the area of ​​the user's eye gaze in order to obtain preset information about the target seen by the user;

[0017] Based on the target preset information and the target preset position corresponding to the target preset information, the target training position selected by the user in the motion space is determined.

[0018] In one embodiment of the present invention, the trajectory motion planning method further includes: after the rehabilitation training is completed, controlling the robotic arm to move to the initial position of the motion space.

[0019] In one embodiment of the present invention, the motion function level corresponding to the reference motion trajectory is one function level higher than the user's current motion function level.

[0020] In one embodiment of the present invention, the trajectory template library for constructing the robotic arm of the rehabilitation training robot includes:

[0021] Acquire multiple trajectory data, including initial trajectory data of each joint of the upper limb when a user with different motor function levels moves from an initial position in the movement space to various training positions in the movement space;

[0022] The multiple initial trajectory data are calculated based on the dynamic time warping algorithm to perform time alignment on the multiple initial trajectory data, thereby obtaining multiple trajectory data.

[0023] Based on the Akaike information criterion and / or the Bayesian information criterion, the number of Gaussian distributions is determined according to multiple trajectory data;

[0024] Construct a Gaussian mixture model containing the number of Gaussian distribution components, and input multiple trajectory data into the Gaussian mixture model for learning to obtain the probability density distribution corresponding to each Gaussian distribution;

[0025] Using the probability density distribution corresponding to each Gaussian distribution as a new feature, multiple trajectory data are input into a Gaussian regression model for learning, in order to extract trajectory templates corresponding to the motor function level.

[0026] In one embodiment of the present invention, the motion space is a nine-square grid.

[0027] As a second aspect of the present invention, the present invention provides a rehabilitation training robot, comprising:

[0028] A controller, comprising a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, it executes the trajectory motion planning method of the rehabilitation training robot as described above.

[0029] A robotic arm that moves from an initial position to a target training position in the motion space according to a reference motion trajectory in order to perform rehabilitation training on the user.

[0030] In one embodiment of the present invention, the rehabilitation training robot further includes:

[0031] Display device, used to display multiple preset information corresponding to cognitive tasks included in rehabilitation training in corresponding preset positions;

[0032] An eye tracker is used to capture the area of ​​a user's eye gaze in order to obtain preset information about the target that the user's eyes see when gazing at the display device.

[0033] This invention provides a trajectory motion planning method for a rehabilitation training robot. First, a trajectory template library for the robotic arm is constructed. This library stores trajectory templates and their corresponding motor function levels. A trajectory template refers to the trajectory of a person at a given motor function level moving from an initial position to a training position in the movement space. Once the user's current motor function level and target training position in the movement space are determined, a reference motion trajectory can be matched in the trajectory template library. The robotic arm is then controlled to move along this reference trajectory to the target training position to complete the user's rehabilitation training. The trajectory template library constructed in this invention includes trajectory templates for different motor function levels and training positions. When performing rehabilitation training on a user, the reference motion trajectory used by the robotic arm is matched according to the user's current motor function level. Therefore, using a matched reference motion trajectory for rehabilitation training reduces the difficulty of rehabilitation training and decreases the risk of secondary injury. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0035] Figure 1 The diagram shown is a schematic flowchart of a trajectory motion planning method for a rehabilitation training robot according to an embodiment of the present invention.

[0036] Figure 2 The diagram shown is a schematic of the user's target training position when the motion space is a nine-square grid according to an embodiment of the present invention.

[0037] Figure 3 The figure shown is a comparison diagram of joint trajectories at different levels of motor function in one embodiment of the present invention.

[0038] Figure 4 The diagram shown is a structural schematic of a robotic arm in a rehabilitation training robot according to an embodiment of the present invention.

[0039] Figure 5 The diagram shown is a schematic flowchart of a trajectory motion planning method for a rehabilitation training robot according to another embodiment of the present invention.

[0040] Figure 6The diagram shown is a schematic flowchart of a trajectory motion planning method for a rehabilitation training robot according to another embodiment of the present invention.

[0041] Figure 7 The diagram shown is a schematic diagram of trajectory data collection in one embodiment of the present invention.

[0042] Figure 8 The diagram shown is a schematic representation of trajectory data without time alignment in one embodiment of the present invention.

[0043] Figure 9 The diagram shown is a schematic representation of trajectory data after time alignment in one embodiment of the present invention.

[0044] Figure 10 The figure shown is an information criterion diagram for the number of Gaussian distributions K in one embodiment of the present invention.

[0045] Figure 11 The diagram shown is a schematic of a Gaussian mixture model learned by the EM algorithm in one embodiment of the present invention.

[0046] Figure 12 The figure shown is a schematic diagram of the Gaussian mixture regression calculation results in one embodiment of the present invention.

[0047] Figure 13 The diagram shown is a block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0048] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] In the description of this invention, it should be noted that the terms "upper", "lower", "front", "horizontal", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0050] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the term "installation" should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral connection; it can refer to a mechanical connection or an electrical connection; it can refer to a direct connection or an indirect connection through an intermediate medium; it can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0051] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings.

[0052] Exemplary control method

[0053] As a first aspect of the present invention, the present invention provides a trajectory motion planning method for a rehabilitation training robot, which controls the robotic arm in the rehabilitation training robot to move along a planned trajectory template to drive the user to perform rehabilitation training. Figure 1 The diagram shown is a schematic flowchart of a trajectory motion planning method for a rehabilitation training robot according to an embodiment of the present invention. Figure 1 As shown, the trajectory motion planning method for a rehabilitation training robot provided by the present invention includes the following steps:

[0054] S1: Construct a trajectory template library for the robotic arm of the rehabilitation training robot. The trajectory template library stores trajectory templates and the corresponding motion function levels of the trajectory templates.

[0055] Specifically, the trajectory template library is pre-built and stores trajectory templates and their corresponding motor function levels. A trajectory template represents the trajectory of a person at a corresponding motor function level moving from an initial position in the movement space to a training position in the movement space.

[0056] Motor function levels are categorized into Brunnstrom Stage 3 and below, Brunnstrom Stage 4, Brunnstrom Stage 5, and Brunnstrom Stage 6 (considered healthy individuals). Brunnstrom Stage 3 and below correspond to the same trajectory template, while other motor function levels correspond to their respective trajectory templates. All trajectory templates in the trajectory template library are labeled, and the label must include at least the motor function level; for example, trajectory template Q. i,j,k,m The trajectory template is labeled with i, j, k, and m, where i represents the motor function level. For example, i=3 indicates a motor function level less than or equal to Brunnstrom stage 3, i=4 indicates a Brunnstrom stage 4, i=5 indicates a Brunnstrom stage 5, and i=6 indicates a Brunnstrom stage 6. j represents the left or right hand. For example, j=1 indicates the left hand, and j=2 indicates the right hand.

[0057] Specifically, the motion space can be a 3x3 grid, such as... Figure 2As shown, the nine-square grid consists of nine squares: square 1 (301), square 2 (302), square 3 (303), square 4 (304), square 5 (305), square 6 (306), square 7 (307), square 8 (308), and square 9 (309). The initial position 313 is set directly below square 6 (306).

[0058] The trajectory template represents the trajectory of the robotic arm from the initial position 313 of the upper limb joint of a person at a corresponding motor function level to a specific cell in the nine-square grid diagram. Specifically, the exact position of each cell can also be represented by a label, such as trajectory template Q. i,j,k,m The corresponding labels for this trajectory template are i, j, k, and m, where k represents the specific position of the grid. For example, K=1 represents the center point of the first grid 301, K=2 represents the center point of the second grid 302, K=3 represents the center point of the third grid 303, K=4 represents the center point of the fourth grid 304, K=5 represents the center point of the fifth grid 305, K=6 represents the center point of the sixth grid 306, K=7 represents the center point of the seventh grid 307, K=8 represents the center point of the eighth grid 308, and K=9 represents the center point of the ninth grid 309.

[0059] For example, trajectory template Q 4,j,1,m This refers to the trajectory template of the upper limb joints of a person with a Brunnstrom stage 4 motor function, from the initial position 313 to the first grid 301 in the nine-square grid diagram.

[0060] It should be noted that the motion space can be the nine-square grid diagram mentioned above, or it can be a six-square grid diagram, a four-square grid diagram, a twelve-square grid diagram, a sixteen-square grid diagram, etc. The present invention does not limit the specific form of the motion space.

[0061] S2: Based on the target training location selected by the user, search the trajectory template library for a reference motion trajectory that matches the user's current motor function level;

[0062] Specifically, the target training location refers to the user's specific location within the motion space, such as... Figure 2 In the nine-square grid, any square can be the user's target training position. For example, the first square 301 can be the target training position for user one, and the second square 302 can also be the target training position for user two.

[0063] Once the user's target training location is determined, a reference motion trajectory matching the user's current motor function level can be found in the trajectory template library based on the target training location.

[0064] For example, the trajectory template library stores trajectory template Q. i,j,k,mWhere i represents the motor function level and k represents the training position in the motion space. Once the user's current motor function level and target training position are determined, a matching reference motion trajectory can be found in the trajectory template library.

[0065] Optionally, when a user is undergoing rehabilitation training, the motor function level corresponding to the reference movement trajectory is one level higher than the user's current motor function level.

[0066] For example, if a user's fitness level is Brunnstrom Stage 3, then the reference fitness track would be track template Q. 4,j,k,m .

[0067] Taking the shoulder flexion / extension angle trajectory template when the right hand moves to the second grid (302) in the nine-square grid as an example, the trajectory template diagrams corresponding to different motor function levels are as follows: Figure 3 As shown, Figure 3 In the diagram, 201 represents the trajectory template Q of the shoulder flexion / extension angle when the right hand of a healthy person (i.e., with a motor function level of Brunnstrom Stage 6) moves to the second square (302) of the nine-square grid. 6,2,2,1 202 indicates a motor function level of Brunnstrom Stage 5. The template Q shows the trajectory of the shoulder joint flexion / extension angle when the right hand moves to the second square (302) in the nine-square grid. 5,2,2,1 203 indicates a Brunnstrom Stage 4 motor function level. The template shows the trajectory of the shoulder flexion / extension angle when the right hand moves to the second square (302) in the nine-square grid. Q4,2,2,1 204 indicates a motor function level of Brunnstrom Stage 3. The template Q shows the trajectory of the shoulder flexion / extension angle when the right hand moves to the second square (302) in the nine-square grid. 3,2,2,1 , Figure 3 The shaded area represents the corresponding variance.

[0068] from Figure 3 As can be seen, the higher the level of motor function, the shorter the time to complete the exercise. A person with a Brunnstrom level 3 needs 2.76 seconds, a Brunnstrom level 4 needs 2.21 seconds, a Brunnstrom level 5 needs 1.75 seconds, and a healthy person only needs 1.03 seconds. If the trajectory template Q for a healthy person is used... 6,2,2,1The rehabilitation training trajectory for a user with a Brunnstrom Stage 3 motor function level requires them to complete the exercise within 1.03 seconds, which necessitates a 62.7% reduction in exercise time – a significant challenge. This invention, however, uses a trajectory template corresponding to a level one level higher than the user's current motor function level as a reference trajectory, employing a progressive rehabilitation method and utilizing the trajectory template Q of a Brunnstrom Stage 4 user. 4,2,2,1 As a reference exercise trajectory used for rehabilitation by users with a Brunnstrom Stage 3 fitness level, users with a Brunnstrom Stage 3 fitness level only need to reduce their exercise time by 20.0%.

[0069] This invention employs progressive rehabilitation training, using joint trajectories at a higher level of motor function as the current user's rehabilitation training trajectory, thereby reducing the difficulty of rehabilitation training and minimizing the risk of secondary injury.

[0070] S3: Control the robotic arm to move to the target training position in the motion space according to the reference motion trajectory, so as to carry out rehabilitation training for the user.

[0071] At the start of training, the robotic arm is in its initial position in the motion space. After the user's reference motion trajectory is found through S2, the user's target hand (e.g., left or right hand) is combined with the robotic arm, and the robotic arm is controlled to move to the target training position in the motion space according to the reference motion trajectory, so as to carry out rehabilitation training for the user.

[0072] This invention provides a trajectory motion planning method for a rehabilitation training robot. First, a trajectory template library for the robotic arm is constructed. This library stores trajectory templates and their corresponding motor function levels. A trajectory template refers to the trajectory of a person at a given motor function level moving from an initial position to a training position in the motion space. Once the user's current motor function level and target training position in the motion space are determined, a reference motion trajectory can be matched in the trajectory template library. The robotic arm is then controlled to move along this reference trajectory to the target training position to complete the user's rehabilitation training. The trajectory template library constructed in this invention includes trajectory templates for different motor function levels and training positions. When performing rehabilitation training on a user, the reference motion trajectory used by the robotic arm is matched according to the user's current motor function level. Therefore, using a matched reference motion trajectory for rehabilitation training reduces the difficulty of rehabilitation training and decreases the risk of secondary injury.

[0073] In one embodiment of the present invention, Figure 4 The diagram shown is a structural schematic of the robotic arm of a rehabilitation training robot according to an embodiment of the present invention. Figure 4As shown, the robotic arm includes multiple degrees of freedom for its moving parts, which correspond one-to-one with the degrees of freedom of the human upper limb joints. Among these degrees of freedom are: shoulder joint external swing or adduction (401), shoulder joint flexion or extension (402), shoulder joint internal rotation or external rotation (403), elbow joint flexion or extension (404), forearm pronation or supination (405), wrist joint ulnar or radial flexion (406), and wrist joint dorsiflexion or palmar flexion (407).

[0074] Correspondingly, the trajectory template includes multiple sub-trajectory templates. Each sub-trajectory template represents the trajectory template of the upper limb joint degrees of freedom parameters during the movement of a person at a corresponding motor function level from an initial position in the movement space to a training position in the movement space. These joint degrees of freedom parameters include joint angles, angular velocities, etc. For example, trajectory template Q... i,j,k,m , i represents the level of motor function, j represents the left or right hand, k represents the training position in the movement space, and m represents the degree of freedom of the movement part. m=1 represents the shoulder joint flexion or extension angle, m=2 represents the shoulder joint abduction or adduction angle, m=3 represents the shoulder joint internal or external rotation angle, m=4 represents the elbow joint flexion or extension angle, m=5 represents the forearm pronation or supination angle, m=6 represents the wrist joint flexion or extension angle, and m=7 represents the wrist joint dorsiflexion or palmar flexion angle.

[0075] For example, trajectory template Q 4,1,4,m This indicates a motor function level of stage 4, with the left hand as the target training hand, and the training position being the trajectory template corresponding to the fourth grid (304) in the nine-square grid diagram. Then, the trajectory template Q... 4,1,4,m It includes 7 sub-trajectory templates, namely Q 4,1,4,1 Q 4,1,4,2 Q 4,1,4,3 Q 4,1,4,4 Q 4,1,4,5 Q 4,1,4,6 Q 4,1,4,7 These are, respectively, the trajectory templates for the shoulder joint flexion or extension angle, shoulder joint abduction or adduction angle, shoulder joint internal rotation or external rotation angle, elbow joint flexion or extension angle, forearm pronation or supination angle, wrist joint flexion or extension angle, and wrist joint dorsiflexion or palmar flexion angle when the robotic arm moves from the initial position to the center point of the fourth grid 304.

[0076] Correspondingly, the reference motion trajectory obtained by matching the user's current motor function level and the target training position includes multiple sub-reference trajectories. S3 (controlling the robotic arm to move to the target training position in the motion space according to the reference motion trajectory) specifically includes the following steps S31-S32:

[0077] S31: Analyze multiple sub-reference trajectories of the reference motion trajectory;

[0078] Specifically, the number of sub-reference trajectory templates for each motion trajectory template can be 7 or less. Therefore, the number of sub-reference trajectories included in the determined reference motion trajectory can be 7 or less. For example, the reference motion trajectory is Q. 4,1,4,m The sub-reference trajectories are Q 4,1,4,1 Q 4,1,4,2 Q 4,1,4,3 Q 4,1,4,4 Q 4,1,4,5 Q 4,1,4,6 Q 4,1,4,7 These correspond to the sub-reference tracks for shoulder flexion or extension, shoulder abduction or adduction, shoulder internal rotation or external rotation, elbow flexion or extension, forearm pronation or supination, wrist flexion or extension, and wrist dorsiflexion or palmar flexion.

[0079] S32: Control the motion of the robotic arm's moving parts corresponding to the sub-reference trajectory according to the sub-reference trajectory.

[0080] Once the sub-reference trajectories included in the reference motion trajectory are analyzed, the corresponding degrees of freedom of the robotic arm's moving parts can be controlled according to the sub-reference trajectories until the rehabilitation training is completed.

[0081] Optionally, when controlling the movement of the robotic arm's moving parts corresponding to the sub-reference trajectory according to the sub-reference trajectory in S32, the following methods can be used:

[0082] (1) Individual motion of the moving parts of the robotic arm:

[0083] The robot arm's moving parts are controlled sequentially according to the sub-reference trajectory, corresponding to the degrees of freedom of the sub-reference trajectory. For example, if the reference trajectory is Q... 4,1,4,m The sub-reference trajectories are Q 4,1,4,1 Q 4,1,4,2 Q 4,1,4,3 Q 4,1,4,4 Q 4,1,4,5 Q 4,1,4,6 Q 4,1,4,7 Then, based on Q... 4,1,4,1 Control the shoulder joint's forward or backward flexion and extension movements. After the exercise, according to Q... 4,1,4,2 Control the shoulder joint's abduction or adduction movements. After the exercise, according to Q... 4,1,4,3 Control the internal or external rotation of the shoulder joint. After the exercise, according to Q... 4,1,4,4 Control the flexion or extension of the elbow joint. After the exercise, according to Q... 4,1,4,5Control the pronation or supination of the forearm; after the movement, according to Q... 4,1,4,6 Control wrist flexion or extension movements; after the exercise, according to Q... 4,1,4,7 Controlling wrist dorsiflexion or palmar flexion movements is used to complete rehabilitation training for the user.

[0084] (2) Combined motion of at least two robotic arm moving parts with degrees of freedom:

[0085] For example, the reference motion trajectory is Q 4,1,4,m The sub-reference trajectories are Q 4,1,4,1 Q 4,1,4,2 Q 4,1,4,3 Q 4,1,4,4 Q 4,1,4,5 Q 4,1,4,6 Q 4,1,4,7 Then, based on Q... 4,1,4,1 Control the flexion or extension of the shoulder joint, while according to Q 4,1,4,2 Control the shoulder joint's abduction or adduction movements. After the exercise, according to Q... 4,1,4,3 Control the internal or external rotation of the shoulder joint, and simultaneously according to Q 4,1,4,4 Control the flexion or extension of the elbow joint. After the exercise, according to Q... 4,1,4,5 Control the pronation or supination of the forearm, and simultaneously according to Q 4,1,4,6 Control wrist flexion or extension movements; after the exercise, according to Q... 4,1,4,7 Controlling wrist dorsiflexion or palmar flexion movements is used to complete rehabilitation training for the user.

[0086] In another embodiment of the invention, such as Figure 5 As shown, the method for determining the user's target training position, i.e., before S2: (based on the target training position selected by the user in the preset information, searching for a reference motion trajectory in the trajectory template library that matches the user's current motor function level), the trajectory motion planning method also includes the following steps S20-S22:

[0087] S20: Obtain the cognitive tasks included in the rehabilitation training, and display the multiple preset information corresponding to the cognitive tasks at the corresponding preset positions on the display device;

[0088] When a user begins rehabilitation training, after the user puts on the robotic arm, a corresponding rehabilitation training scenario is generated. The training scenario includes a set of cognitive tasks, such as the cognitive N-Back task. For each cognitive task, preset display information (such as the answer to the cognitive task) will randomly appear at a preset position on the display device.

[0089] The preset positions can be random or pre-designed. For example, the preset positions for the preset display information corresponding to cognitive task 1 are the first grid 301 and the ninth grid 309.

[0090] Taking the motion space as a nine-square grid as an example, combined with Figure 2 As shown, the preset display information for each cognitive task is the first answer 310 and the second answer 311, and the first answer 310 and the second answer 311 are randomly displayed in any two of the nine grids in the nine-square grid diagram. Figure 2 As shown, for example, the first answer 310 is displayed in the seventh square 307, and the second answer 311 is displayed in the third square 303.

[0091] S21: Control the eye tracker to capture the area of ​​the user's eye gaze in order to obtain preset information about the target seen by the user;

[0092] After the preset display information corresponding to the cognitive task is displayed on the nine-square grid, the user looks at the nine-square grid and controls the eye tracker to capture the user's gaze area in order to obtain the target preset information seen by the user. That is, the eye tracker determines whether the user is looking at the first answer 310 or the second answer 311. If the user is looking at the first answer 310, then the first answer 310 is the target preset information seen by the user.

[0093] S22: Determine the target training location selected by the user in the motion space based on the target preset information and the target preset location corresponding to the target preset information.

[0094] Once the target information for the user's gaze is determined, the target training location selected by the user can be determined based on the target information and the corresponding preset location. For example, combining... Figure 2 As shown, if the user gazes at the first answer 310 and then at position 312, then the seventh grid 307, where the first answer 310 is located, can be determined as the target training position. Once the target training position is determined, the reference motion trajectory can be determined based on it.

[0095] In this invention, a cognitive N-Back task is designed in the rehabilitation training scenario, which trains the user's cognitive ability while training the user's limbs, and can improve the brain's working memory ability while the limbs are being rehabilitated.

[0096] Optionally, after the rehabilitation training is completed, the robotic arm is controlled to move to the initial position in the motion space.

[0097] In another embodiment of the invention, such as Figure 6 As shown, the specific construction method for building the trajectory template library, namely S1 (building the trajectory template library for the robotic arm of the rehabilitation training robot), includes the following steps:

[0098] S10: Acquire multiple trajectory data, including the initial trajectory data of each joint of the upper limb when a person with different motor function levels moves from the initial position in the movement space to various training positions in the movement space;

[0099] For example, multiple initial trajectory data include multiple sets of initial trajectory data with a Brunnstrom stage 3 motor function level, multiple sets of initial trajectory data with a Brunnstrom stage 4 motor function level, multiple sets of initial trajectory data with a Brunnstrom stage 5 motor function level, and multiple sets of initial trajectory data with a Brunnstrom stage 6 motor function level.

[0100] The number of trajectories in the multiple sets of initial trajectory data corresponding to each motor function level can be the same or different.

[0101] Initial trajectory data refers to the initial trajectory data of each degree of freedom of the upper limbs of a person with a certain level of motor function when moving from the initial position in the movement space to the training position in the movement space.

[0102] Combination Figure 7 As shown, the initial trajectory data can be obtained in the following ways:

[0103] (1) Nine balls (namely, ball 101, ball 102, ball 103, ball 104, ball 105, ball 106, ball 107, ball 108, and ball 109) are suspended at a distance of L1 = 0.3m in front of a person. The fifth ball 105, located in the center, is aligned with the C7 cervical vertebra of the human body. The horizontal distance between two adjacent balls is L2 = 0.3m, and the vertical distance is L3 = 0.3m.

[0104] (2) The person’s left and right palms touch 9 balls respectively, and the angles of each degree of freedom of the upper limbs are collected when touching each ball. The initial trajectory data includes the angles of each degree of freedom.

[0105] S11: Calculate multiple initial trajectory data based on the dynamic time warping algorithm to align the multiple initial trajectory data in time and obtain multiple trajectory data.

[0106] Taking a nine-square grid as an example, since different people take different amounts of time to complete the movement from the initial position to the training position, the length of the collected data is different. Therefore, it is necessary to first align the initial trajectory data in time.

[0107] This invention uses the Dynamic Time Warping (DTW) algorithm to solve the time alignment problem of nine-square grid motion data. The DTW algorithm finds the optimal alignment path between two time series by aligning their time axes to minimize their cumulative distance, and can handle non-linear variations on the time axis. Specifically, it allows certain points on the time axis to be repeatedly matched (i.e., "stretching" or "compressing" the time axis); and finds an optimal matching path that minimizes the sum of the distances between corresponding points in the two sequences.

[0108] Specifically, the DTW algorithm includes the following steps:

[0109] Suppose we have two time series: X=[x1,x2,…,xn], Y=[y1,y2,…,ym]

[0110] Step 1: Construct a distance matrix. Create an n×m matrix D, where each element D(i,j) represents the distance between xi and yj (usually using Euclidean distance or other distance metrics): D(i,j) = |xi yj∣

[0111] Step 2: Initialize the cumulative distance matrix. Construct an n×m cumulative distance matrix C, where C(i,j) represents the minimum cumulative distance from the starting point (1,1) to the point (i,j).

[0112] Initialize the first point: C(1,1) = D(1,1), initialize the first row and first column: C(i,1) = C(i,1) 1,1)+D(i,1) (first column), C(1,j)=C(1,j) 1)+D(1,j) (first row)

[0113] Step 3: Recursively calculate the cumulative distance. For other points (i,j) in the matrix, calculate their cumulative distance:

[0114] C(i,j)=D(i,j)+min (C(i 1,j),C(i,j 1),C(i 1,j 1))

[0115] The minimum value here represents the optimal path selected from three possible paths.

[0116] Step 4: Backtrack the optimal path. Starting from the endpoint (n,m), backtrack to the starting point (1,1) to find the path with the minimum cumulative distance. Backtracking rule: Select the adjacent point (left, bottom, or bottom left) that minimizes the cumulative distance.

[0117] Step 5: Calculate the DTW distance, which is the value of the cumulative distance matrix C(n,m), representing the minimum cumulative distance between two time series.

[0118] Taking the initial trajectory data of the shoulder joint flexion / extension angles when the right hand of a person with a motor function level of Brunnstrom Stage 5 moves from the initial position to the center point of the second grid 302 in the nine-square grid as an example, time alignment was performed. Figure 8 This is a schematic diagram of misaligned trajectory data, from Figure 8 The data shows that the lengths of the 54 initial trajectory data sets are all different, with the shortest data being only 0.8 seconds and the longest data reaching 3.3 seconds. Figure 9 This is a trajectory data map after aligning 54 sets of initial trajectory data using the DTW algorithm. Figure 9 As can be seen, the lengths of the 54 sets of trajectory data are consistent, and the data pattern shows a clear single-peak curve.

[0119] S12: Determine the number of Gaussian distributions based on multiple trajectory data, using the Akaike information criterion and / or the Bayesian information criterion.

[0120] Before modeling a Gaussian mixture model, the number of Gaussian distributions in the Gaussian mixture model must first be determined.

[0121] Specifically, the goal of the Akaike Information Criterion (AIC) is to select a model that maximizes the ability to predict new data. The formula for AIC is: Where: L is the maximum likelihood of the model (i.e., the goodness of fit of the model to the data), k is the number of parameters in the model (i.e., the complexity of the model), and ln(L) is the log-likelihood of the model. The first term of the formula... The AIC (Alignment and Compatibility Index) is used to measure the goodness of fit of the model. The better the fit, the larger the log-likelihood value, and the smaller this term. The second term, 2k, penalizes the complexity of the model; the more parameters and the more complex the model, the larger this term. Choosing the model with the lowest AIC value achieves a balance between goodness of fit and complexity.

[0122] The Bayesian Information Criterion (BIC) aims to select a model that is most likely to generate data from a Bayesian perspective. The formula for BIC is: Where: L is the maximum likelihood of the model, k is the number of parameters in the model, n is the sample size, and ln(n) is the natural logarithm of the sample size. The first term of the formula... Similar to AIC, it is used to measure the goodness of fit of the model. The second term of the formula... It is used to penalize the complexity of the model, but the penalty is stronger than AIC (especially when the sample size n is large). Choosing the model with the smallest BIC value can achieve a more rigorous balance between goodness of fit and complexity.

[0123] Taking the trajectory data of the shoulder joint flexion / extension angles of 54 Brunnstrom Phase 5 participants as their right hand moved from the initial position to the center point of the second grid 302 in the nine-square grid as an example Figure 10 The diagram shown is an information criterion diagram for the Gaussian distribution quantity K according to the present invention. Figure 10 The solid line represents the AIC criterion, and the dashed line represents the BIC criterion. To determine the distribution quantity K, K was increased sequentially from 1 to 20, and the AIC and BIC values ​​of the trajectory data of the shoulder joint flexion / extension angles of 54 groups of Brunnstrom Phase 5 individuals as their right hand moved from the initial position to the center point of the second grid 302 in the nine-square grid were calculated for different K values. The results are shown in the figure. It can be seen from the figure that when K < 6, AIC and BIC decrease rapidly, and when K > 6, AIC and BIC tend to remain constant. Therefore, K = 6 was chosen as the Gaussian distribution quantity for the Gaussian mixture model.

[0124] S13: Construct a Gaussian mixture model containing a number of Gaussian distribution components, and input multiple trajectory data into the Gaussian mixture model for learning to obtain the probability density distribution corresponding to each Gaussian distribution;

[0125] Taking the trajectory data of the shoulder flexion / extension angles of 54 groups of Brunnstrom Phase 5 participants as their right hand moves from the initial position to the center point of the second grid (302) in a nine-square grid as an example, after determining the number of Gaussian distributions, a Gaussian mixture model containing the number of Gaussian distribution components can be constructed. The trajectory data of the shoulder flexion / extension angles of the 54 groups of Brunnstrom Phase 5 participants as their right hand moves from the initial position to the center point of the second grid (302) in a nine-square grid is input into the Gaussian mixture model for learning. The main purpose of model learning is to determine the parameters of the Gaussian mixture model, which include the weights of each Gaussian distribution. mean Covariance Matrix The parameters of the Gaussian mixture model are estimated using the Expectation-Maximization (EM) algorithm.

[0126] Specifically, the method for estimating the parameters of a Gaussian mixture model using the EM algorithm includes the following steps:

[0127] (1) Initialize the parameters of the Gaussian mixture model:

[0128] Randomly initialize weights mean Covariance Matrix .

[0129] (2) Expectation Step:

[0130] Calculate each data point The posterior probability (responsibility value) of belonging to the k-th Gaussian distribution:

[0131] ,in Representing data points The probability of belonging to the k-th component.

[0132] (3) M-step (Maximization Step):

[0133] Update the parameters of the Gaussian mixture model:

[0134]

[0135]

[0136]

[0137] Where N is the total number of data points.

[0138] (4) Iteration: Repeat the E-step and M-step until the parameters converge or the maximum number of iterations is reached.

[0139] Figure 11 The diagram shows a Gaussian mixture model learned by the EM algorithm according to the present invention. The black dashed line in the diagram represents the trajectory data of the shoulder joint flexion / extension angle when the right hand of 54 Brunnstrom Phase 5 people moves from the initial position to the center point of the second grid 302 in the nine-grid diagram. The green ellipse represents the Gaussian distribution. One confidence ellipse corresponds to one Gaussian distribution. There are a total of 6 confidence ellipses.

[0140] Each confidence ellipse corresponds to a probability density distribution of a Gaussian distribution.

[0141] S14: Using the probability density distribution corresponding to each Gaussian distribution as a new feature, input multiple trajectory data into the Gaussian regression model for learning, in order to extract the trajectory template corresponding to the motor function level.

[0142] Taking the trajectory data of the shoulder joint flexion / extension angles as the right hand of 54 Brunnstrom Phase 5 participants moved from the initial position to the center point of the second grid 302 in the nine-square grid as an example:

[0143] After determining the probability density distribution using a Gaussian mixture model, a trajectory template is generated based on the Gaussian mixture regression algorithm. Gaussian Mixture Regression (GMR) is a regression method based on a Gaussian mixture model that uses conditional probability to make predictions by modeling the joint probability distribution of input and output variables.

[0144] Specifically, the steps of the Gaussian mixture regression algorithm are as follows:

[0145] (1) The probability density function of the Gaussian mixture model is known to be:

[0146]

[0147] (2) Assuming the input variable is x and the output variable is y, then the joint variable z = [x ,y ] The joint probability distribution can be expressed as:

[0148]

[0149] The mean and covariance matrix of the k-th Gaussian distribution can be represented in blocks as follows:

[0150]

[0151] (3) Given input x, the conditional distribution of output y is still in Gaussian mixture form:

[0152]

[0153] in:

[0154] The weights of the k-th Gaussian distribution are calculated using the following formula:

[0155]

[0156] It is the conditional mean, and the calculation formula is:

[0157]

[0158] It is the conditional covariance matrix, and its calculation formula is:

[0159]

[0160] (4) Gaussian mixture regression: For a given input x, the predicted value of the output y can be calculated using the expected value of the conditional distribution.

[0161]

[0162] The uncertainty of prediction can be expressed through the conditional covariance matrix. To describe.

[0163] Given a Gaussian mixture model, Gaussian mixture regression only requires calculating the conditional mean of each Gaussian distribution based on the given input x (original data) and the formula in step (3). and weight Then, calculate the expected value according to the formula in step (4). This allows us to obtain a trajectory template of the shoulder joint flexion / extension angle as the user's right hand moves from the initial position to the center point of the second grid 302 in the nine-grid diagram. Figure 12 This is a schematic diagram of a Gaussian mixture regression calculation result incorporating the present invention. Figure 12 The solid black line in the middle represents the Gaussian mixture regression expectation value Q of the trajectory data of the shoulder joint flexion / extension angles as the right hand of 54 Brunnstrom Phase 5 participants moved from the initial position to the center point of the second grid (302) in the nine-square grid. 5,2,2,1 The shaded area represents the standard deviation of the Gaussian mixture regression.

[0164] The expected values ​​obtained from all Gaussian mixture regressions are stored as trajectory templates Qi,j,k,m for use by the robotic arm during user training. Here, i=3 to 6, where i=3 represents a person with Brunnstrom stage 3 motor function, i=4 represents a person with Brunnstrom stage 4, i=5 represents a person with Brunnstrom stage 5, and i=6 represents a person with Brunnstrom stage 6. For a person in phase 6; j=1~2, j=1 represents the left hand, j=2 represents the right hand; k=1~9, k=1 represents moving to the center point of the first square 301 of the 3x3 grid, k=2 represents moving to the center point of the second square 302 of the 3x3 grid, k=3 represents moving to the center point of the third square 303 of the 3x3 grid, k=4 represents moving to the center point of the fourth square 304 of the 3x3 grid, k=5 represents moving to the center point of the fifth square 305 of the 3x3 grid, k=6 represents moving to the center point of the sixth square 306 of the 3x3 grid, k=7 represents... The movement is directed to the center point of the seventh grid (307) in the 3x3 grid diagram. k=8 indicates the movement is directed to the center point of the eighth grid (308) in the 3x3 grid diagram. k=9 indicates the movement is directed to the center point of the ninth grid (309) in the 3x9 grid diagram. m=1 to 7, where m=1 represents the shoulder joint flexion / extension angle, m=2 represents the shoulder joint abduction / adduction angle, m=3 represents the shoulder joint internal / external rotation angle, m=4 represents the elbow joint flexion / extension angle, m=5 represents the forearm pronation / supination angle, m=6 represents the wrist joint flexion / extension angle, and m=7 represents the wrist joint dorsiflexion / palmar flexion angle.

[0165] Exemplary robot

[0166] As a second aspect of the present invention, the present invention also provides a rehabilitation training robot, comprising:

[0167] The controller includes a processor and a memory. The memory stores a computer program. When the processor executes the computer program, it executes the trajectory motion planning method for the rehabilitation training robot described above.

[0168] A robotic arm moves from an initial position along a reference motion trajectory to a target training position in the motion space to perform rehabilitation training on the user.

[0169] The rehabilitation training robot provided by this invention first constructs a trajectory template library for the robotic arm. This library stores trajectory templates and their corresponding motor function levels. A trajectory template refers to the trajectory of a person at a given motor function level moving from an initial position to a training position in the movement space. Once the user's current motor function level and target training position in the movement space are determined, a reference motion trajectory can be matched in the trajectory template library. The robotic arm is then controlled to move along this reference trajectory to the target training position to complete the user's rehabilitation training. The trajectory template library constructed by this invention includes trajectory templates for different motor function levels and different training positions. When performing rehabilitation training on a user, the reference motion trajectory used by the robotic arm is matched according to the user's current motor function level. Therefore, using a matched reference motion trajectory for rehabilitation training reduces the difficulty of rehabilitation training and decreases the risk of secondary injury.

[0170] Optional, rehabilitation training robots may also include:

[0171] Display device, used to display multiple preset information corresponding to cognitive tasks included in rehabilitation training in corresponding preset positions;

[0172] An eye tracker is used to capture the area of ​​a user's eye gaze in order to obtain preset information about the target that the user's eyes see when they are looking at a display device.

[0173] Exemplary electronic devices

[0174] As a third aspect of the present invention, the present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, it executes the trajectory motion planning method for the rehabilitation training robot described above.

[0175] Specifically, the internal structure of electronic devices can be as follows: Figure 13As shown, the electronic device includes a processor, a memory, a network interface, and an input device connected via a device bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores operating devices and computer programs. The internal memory provides an environment for the operation of the operating devices and computer programs stored in the non-volatile storage medium. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it follows the steps of the trajectory motion planning method for a rehabilitation training robot according to various embodiments of this specification, as described in the above embodiments.

[0176] The processor may include the main processor, as well as baseband chips, modems, etc.

[0177] The memory stores a program that executes the technical solution of this invention, and may also store operating devices and other key business functions. Specifically, the program may include program code, which includes computer operation instructions. More specifically, the memory may include read-only memory (ROM), other types of static storage devices capable of storing static information and instructions, random access memory (RAM), other types of dynamic storage devices capable of storing information and instructions, disk storage, flash memory, etc.

[0178] The processor can be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present invention. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0179] Input devices may include devices that receive data and information input by the user, such as keyboards, mice, cameras, scanners, light pens, voice input devices, touch screens, pedometers, or gravity sensors.

[0180] Output devices may include devices that allow information to be output to the user, such as displays, printers, speakers, etc.

[0181] The communication interface may include any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0182] The processor executes the program stored in the memory and calls other devices, which can be used to implement the various steps of the trajectory motion planning method for any rehabilitation training robot provided in the above embodiments of this specification.

[0183] The electronic device may also include a display component and a voice component. The display component may be a liquid crystal display screen or an e-ink display screen. The input device of the electronic device may be a touch layer covering the display component, or a button, trackball or touchpad set on the casing of the electronic device, or an external keyboard, touchpad or mouse, etc.

[0184] Those skilled in the art will understand that Figure 13 The structures shown are merely block diagrams of a portion of the structure related to the scheme described in this specification, and do not constitute a limitation on the electronic devices to which the scheme described in this specification is applied. Specific electronic devices may include more or fewer components than those shown in the figures, or may combine certain components, or may have different component arrangements.

[0185] Exemplary computer program products and storage media

[0186] In addition to the methods and devices described above, the trajectory motion planning method for rehabilitation training robots provided in the embodiments of this specification can also be a computer program product, which includes computer program instructions that, when executed by a processor, cause the processor to perform the steps in the trajectory motion planning method for rehabilitation training robots according to various embodiments of this specification as described in the "Exemplary Methods" section above.

[0187] The aforementioned computer program product can be implemented through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium; in another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0188] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this specification. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages.

[0189] Furthermore, embodiments of this specification also provide a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor of the steps in the trajectory motion planning method for a rehabilitation training robot according to various embodiments of this specification as described in the "Exemplary Methods" section above.

[0190] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this specification can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0191] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0192] The embodiments described above are merely illustrative of several implementation methods outlined in this specification. While the descriptions are specific and detailed, they should not be construed as limiting the scope of the solutions provided in this specification. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this specification, and these all fall within the scope of protection of this specification. Therefore, the scope of protection for this patent should be determined by the appended claims.

Claims

1. A trajectory motion planning method for a rehabilitation training robot, characterized in that, include: A trajectory template library for the robotic arm of a rehabilitation training robot is constructed. The trajectory template library stores trajectory templates and the corresponding motor function levels of the trajectory templates. The trajectory templates represent the trajectory of a person of the corresponding motor function level moving from an initial position in the movement space to a training position in the movement space. Based on the target training location selected by the user, a reference motion trajectory matching the user's current motor function level is searched in the trajectory template library; The robotic arm is controlled to move to the target training position in the motion space according to the reference motion trajectory in order to perform rehabilitation training on the user; The trajectory template library for constructing the robotic arm of the rehabilitation training robot includes: Acquire multiple initial trajectory data, which include initial trajectory data of each joint of the upper limb when a user with different motor function levels moves from an initial position in the movement space to various training positions in the movement space; The multiple initial trajectory data are calculated based on the dynamic time warping algorithm to perform time alignment on the multiple initial trajectory data, thereby obtaining multiple trajectory data. Based on the Akaike information criterion and / or the Bayesian information criterion, the number of Gaussian distributions is determined according to multiple trajectory data; Construct a Gaussian mixture model containing the number of Gaussian distribution components of the Gaussian distribution, and input multiple trajectory data into the Gaussian mixture model for learning to obtain the probability density distribution corresponding to each Gaussian distribution; Using the probability density distribution corresponding to each Gaussian distribution as a new feature, multiple trajectory data are input into a Gaussian regression model for learning, in order to extract trajectory templates corresponding to the motor function level.

2. The trajectory motion planning method according to claim 1, characterized in that, The robotic arm includes multiple degrees of freedom for its moving parts, which correspond one-to-one with the degrees of freedom of the human upper limb joints. These include the shoulder joint's external swing or adduction, shoulder joint's flexion or extension, shoulder joint's internal or external rotation, elbow joint's flexion or extension, forearm's pronation or supination, wrist joint's ulnar or radial flexion, and wrist joint's dorsiflexion or palmar flexion. The trajectory template includes multiple sub-trajectory templates. The sub-trajectory templates are time series of upper limb joint degree-of-freedom parameters during the process of a person of a corresponding motor function level moving from an initial position in the movement space to a training position in the movement space. The joint degree-of-freedom parameters include joint angles and angular velocities.

3. The trajectory motion planning method according to claim 2, characterized in that, The step of controlling the robotic arm to move to the target training position in the motion space according to the reference motion trajectory includes: Analyze multiple sub-reference trajectories of the reference motion trajectory; The movement of the robotic arm's moving parts corresponding to the sub-reference trajectory is controlled according to the sub-reference trajectory.

4. The trajectory motion planning method according to claim 1, characterized in that, Before searching for a reference motion trajectory in the trajectory template library that matches the user's current motor function level based on the target training location selected by the user, the trajectory motion planning method further includes: Acquire the cognitive tasks included in the rehabilitation training, and display multiple preset information corresponding to the cognitive tasks at corresponding preset positions on the display device; The eye tracker is controlled to capture the area of ​​the user's eye gaze in order to obtain preset information about the target seen by the user; Based on the target preset information and the target preset position corresponding to the target preset information, the target training position selected by the user in the motion space is determined.

5. The trajectory motion planning method according to claim 4, characterized in that, Also includes: Once the rehabilitation training is completed, the robotic arm is controlled to move to the initial position in the motion space.

6. The trajectory motion planning method according to claim 1, characterized in that, The motion function level corresponding to the reference motion trajectory is one function level higher than the user's current motion function level.

7. The trajectory motion planning method according to claim 1, characterized in that, The motion space is a nine-square grid.

8. A rehabilitation training robot, characterized in that, include: A controller, comprising a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, it executes the trajectory motion planning method of the rehabilitation training robot according to any one of claims 1-7; A robotic arm that moves from an initial position to a target training position in the motion space according to a reference motion trajectory in order to perform rehabilitation training on the user.

9. The rehabilitation training robot according to claim 8, characterized in that, Also includes: Display device, used to display multiple preset information corresponding to cognitive tasks included in rehabilitation training in corresponding preset positions; An eye tracker is used to capture the area of ​​a user's eye gaze in order to obtain preset information about the target that the user's eyes see when gazing at the display device.

Images