Mirror compliant control method for lower limb rehabilitation robot based on dynamic motion primitives

Abstract

The application relates to a kind of lower limb rehabilitation robot mirror image compliant control methods based on dynamic motion primitives, comprising: step 1: collecting the robot end trajectory data of the same side of the healthy leg of the patient during the flexion and extension training process of the patient, filtering the data and saving to the computer end;Step 2: according to the data after filtering, using dynamic motion primitives method, combined with the lead model, the motion trajectory model of lower limb rehabilitation robot is established;Step 3: design trajectory tracking controller, accurately track the motion trajectory of the affected leg;Step 4: design adaptive lead controller, correct the input trajectory of trajectory tracking controller;Step 5: combine trajectory tracking controller with adaptive lead controller, design hierarchical control strategy.The application can effectively improve the coordination of the patient's legs by mirror image rehabilitation for the patient;The safety and comfort of the patient during the rehabilitation training process can be effectively improved by compliant control for the patient's affected leg.

Description

Technical Field

[0001] This invention relates to the field of rehabilitation robot control technology, specifically to a mirror compliance control method for a lower limb rehabilitation robot based on dynamic motion primitives. Background Technology

[0002] As the aging population becomes increasingly prominent, the number of patients with lower limb dysfunction such as hemiplegia caused by diseases is also increasing. In response to the growing demand for rehabilitation training services, rehabilitation robot systems are being used in the rehabilitation training of clinical patients, effectively improving the limb motor function of stroke hemiplegic patients.

[0003] Robot-assisted training, as an effective method for stroke rehabilitation, involves a robot driving the patient's injured limb to perform corresponding movements. However, in this purely passive training mode, the patient cannot change the robot's movement, especially when they feel uncomfortable or even in pain. For patients with hemiplegia, robot mirror therapy is a very effective rehabilitation method. By having the affected leg imitate the movement of the healthy leg, it increases the patient's motivation for rehabilitation training. Simultaneously, both legs participate in rehabilitation training, improving the coordination of lower limb movements. Dynamic motion primitives are a representative method for modeling robot motion trajectories through imitation learning. Especially for mirror rehabilitation, the imitation characteristics of dynamic motion primitives perfectly meet the patient's behavioral requirement of the affected leg imitating the healthy leg. However, as rehabilitation training progresses and the patient's muscle strength gradually increases, the affected leg regains some mobility. The passive tracking of the healthy leg's movement by the affected leg lacks sufficient flexibility. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by providing a mirror-compliant control method for lower limb rehabilitation robots based on dynamic motion primitives, thereby improving the comfort of patients' rehabilitation training.

[0005] This invention is achieved through the following technical solution:

[0006] A mirror-compliant control method for lower limb rehabilitation robots based on dynamic motion primitives includes the following steps:

[0007] Step 1: Based on the position of the patient's healthy leg, install the detachable pressure sensor on the lower leg link on the same side as the patient's healthy leg, collect the robot end-effector trajectory data on the same side as the patient's healthy leg during the patient's flexion and extension training, filter the data and save it to the computer.

[0008] Step 2: Based on the filtered data, use the dynamic motion primitive method combined with the admittance model to establish a motion trajectory model for the lower limb rehabilitation robot. Based on the interaction force between the healthy leg and the robot, plan the motion trajectory of the patient's affected leg.

[0009] Step 3: Design a trajectory tracking controller to accurately track the movement trajectory of the affected leg;

[0010] Step 4: Design an adaptive admittance controller and use the adaptive admittance method to correct the input trajectory of the trajectory tracking controller;

[0011] Step 5: Combine the trajectory tracking controller with the adaptive admittance controller to design a hierarchical control strategy and define the human-computer interaction force threshold. Select different control strategies according to different interaction forces. If the force is within the threshold range, use the trajectory tracking controller to complete the training task; if the force is not within the threshold range, return to step 4 to activate the adaptive admittance controller and correct the input trajectory of the trajectory tracking controller to achieve compliant control of the patient's affected leg.

[0012] Furthermore, the specific steps of step 2 are as follows:

[0013] Step 2.1: Establish a motion trajectory model based on the robot end effector trajectory data obtained in Step 1. The mathematical formula for the motion trajectory model is:

[0014]

[0015] Where, x, It refers to the robot's trajectory, velocity, and acceleration, α. x and β x It is a proportionality constant and β x =α x / 4;τ s The duration of motion is represented by τ. s Adjust the convergence speed of the trajectory; g is the target position of the trajectory, which changes the amplitude and frequency of the trajectory; f(x) is used to adjust the endpoint and shape of the trajectory, specifically:

[0016]

[0017] in It is the Gaussian function, N represents the number of Gaussian functions, ω i These are the weights of each basis function that determine the shape of the gait trajectory;

[0018] Step 2.2: Add a human-computer interaction term to the motion trajectory model in Step 2.1, and describe the human-computer interaction using an admittance model. The admittance model is as follows:

[0019]

[0020] Where F is the interaction force between humans and machines, and k h and d h These are impedance model parameters, x and It is the displacement and acceleration output by the admittance model based on the interaction force between the human and the machine;

[0021] Step 2.3: Let the reference trajectory x of the admittance model in Step 2.2 be... d =0, adjust the output displacement x of the admittance model according to the human-machine interaction force F, and let the target position g of the trajectory model in step 2.1 be x, so as to adjust the target position of the trajectory model according to the human-machine interaction force and plan the movement trajectory of the patient's affected leg.

[0022] Furthermore, the specific steps for step 3 are as follows:

[0023] Step 3.1: Establish the dynamic model of the lower limb rehabilitation robot, specifically as follows:

[0024]

[0025] Where, θ, and These are the hip joint angle, angular velocity, and angular acceleration of the rehabilitation robot, respectively. Furthermore, M(θ) is the inertia matrix. Here, G(θ) is the Coriolis and centrifugal force matrix, G(θ) is the robot's gravity matrix, and τ is the control torque applied to the joints. hr It is the torque for human-computer interaction.

[0026] Let x1 = θ, u = τ, the dynamic equations of the lower limb rehabilitation robot are rewritten as state-space equations:

[0027]

[0028] in,

[0029] f(x1,x2)=-M -1 (x1)(C(x1,x2)+G(x1))

[0030] M(x1)=(m1+m2)l 2 / 3

[0031] C(x1,x2)=-2m2l 2 sinx1cosx1

[0032] G(x1)=((m1+m2)glcosx1) / 2

[0033] m1 and m2 are the masses of the thigh and lower leg connecting rods, and l is the length of the thigh and lower leg connecting rods.

[0034] Step 3.2: Design a trajectory tracking controller and compensate for the human-computer interaction force. The estimated value of the human-computer interaction force is:

[0035]

[0036] The trajectory tracking controller is:

[0037]

[0038] Furthermore, the specific steps of step 4 are as follows:

[0039] Step 4.1: Establish the admittance model, specifically as follows:

[0040]

[0041] Where ΔT is the change in the interaction torque between the human and the robot, M, B, and K are the inertia, damping, and stiffness of the admittance model, respectively, and θ, and These are the robot's angular displacement, angular velocity, and angular acceleration, respectively.

[0042] Step 4.2: Design the adaptive law for admittance parameters, specifically as follows:

[0043]

[0044]

[0045] Where B0 and K0 are base values, a j and b j (j=1,2,3) are the coefficients for adjusting the damping and stiffness of the robot admittance model, respectively, and B lj and K lj (j=1,2) are the upper and lower thresholds for damping and stiffness, and T is the interaction torque between the human body and the robot.

[0046] Furthermore, the specific steps of step 5 are as follows:

[0047] Step 5.1: Design a hierarchical control strategy, including an upper-level adaptive admittance controller and a lower-level position tracking controller. When the human interacts with the robot, the upper-level admittance controller adjusts the motion trajectory of the lower limb rehabilitation robot and transmits the adjustment amount as input to the lower-level position tracking controller to control the robot to drive the patient's affected leg to track the trajectory.

[0048] Step 5.2: Define the range of plantar pressure based on the interaction between the human and the robot. Divide the interaction area between the patient and the robot into regions according to the different magnitudes of the interaction force between the human and the robot, and execute the corresponding control strategy according to the corresponding region.

[0049] Furthermore, the areas of interaction between the patient and the robot are divided into: passive motion range, cooperative motion range, and abnormal motion range.

[0050] When the human-computer interaction force is within the passive motion range, the robot performs trajectory tracking tasks;

[0051] When the human-computer interaction force is within the range of collaborative motion, the patient can actively apply force to change the robot's motion state;

[0052] When the human-computer interaction force is within an abnormal range of motion, the robot stops moving.

[0053] A mirror-compliant control device for a lower limb rehabilitation robot based on dynamic motion primitives, including

[0054] The filtering module is used to collect the robot end-effector trajectory data on the same side as the patient's healthy leg during flexion and extension training, filter the data, and save it to the computer.

[0055] The trajectory generation module is used to establish a motion trajectory model of the lower limb rehabilitation robot based on the filtered data, using the dynamic motion primitive method combined with the admittance model, and to plan the motion trajectory of the patient's affected leg based on the interaction force between the healthy leg and the robot.

[0056] The policy control module includes:

[0057] A trajectory tracking controller is used to accurately track the movement trajectory of the affected leg;

[0058] An adaptive admittance controller is used to correct the input trajectory of the trajectory tracking controller;

[0059] By combining a trajectory tracking controller with an adaptive admittance controller, a hierarchical control strategy is designed, and a human-machine interaction force threshold is defined. Different control strategies are selected based on different interaction forces.

[0060] Moreover, the trajectory generation module includes

[0061] The input module is used to build a motion trajectory model based on the robot's end effector trajectory data;

[0062] The computation module is used to add human-computer interaction terms to the motion trajectory model and describe the human-computer interaction using an admittance model.

[0063] The planning module sets initial parameters and the target position of the trajectory model, adjusts the target position of the trajectory model based on human-computer interaction force, and plans the movement trajectory of the patient's affected leg.

[0064] An electronic device includes a memory in which processors are interconnected, the memory storing computer instructions, and the processors executing the computer instructions to perform the mirror compliance control method for a lower limb rehabilitation robot based on dynamic motion primitives as described above.

[0065] A computer storage medium storing computer instructions, the computer instructions being configured to cause a computer to execute the mirror compliant control method for a lower limb rehabilitation robot based on dynamic motion primitives as described above.

[0066] The advantages and beneficial effects of this invention are as follows:

[0067] This invention is applicable to mirror rehabilitation for patients with unilateral lower limb disability. By simply modifying a lower limb rehabilitation robot, the hardware requirements for a mirror rehabilitation strategy can be met. Mirror rehabilitation can effectively improve the coordination of the patient's two legs; compliant control of the affected leg can effectively prevent secondary injury during training, improving the safety and comfort of the patient during rehabilitation training. Attached Figure Description

[0068] Figure 1 This is a flowchart of a mirror-compliant control method for a lower limb rehabilitation robot based on dynamic motion primitives.

[0069] Figure 2 This is a schematic diagram of the structure of a mirrored lower limb rehabilitation robot.

[0070] Figure 3 This is a structural diagram of the mirrored rehabilitation strategy for a lower limb rehabilitation robot.

[0071] Figure 4 A diagram illustrating the division of human-computer interaction forces.

[0072] For those skilled in the art, other related figures can be obtained from the above figures without any creative effort. Detailed Implementation

[0073] To enable those skilled in the art to better understand the present invention, the technical solution of the present invention will be further described below with reference to specific embodiments.

[0074] A mirror-compliant control method for lower limb rehabilitation robots based on dynamic motion primitives, such as Figure 1 As shown, this method includes the following steps:

[0075] Step 1: Based on the position of the patient's healthy leg, install a detachable pressure sensor on the lower leg link on the same side as the patient's healthy leg, collect the robot end-effector trajectory data on the same side as the patient's healthy leg during flexion and extension training, filter the data and save it to the computer.

[0076] In this embodiment, as follows Figure 2The robot shown is a lower limb rehabilitation robot. Foot pressure sensors are installed on the foot pedals on both sides of the robot, and detachable pressure sensors are installed on the lower leg connecting rod on the same side as the patient's healthy leg.

[0077] Step 2: Based on the processed data, using the dynamic motion primitive method combined with an admittance model, establish a motion trajectory model for the lower limb rehabilitation robot. Based on the interaction force between the healthy leg and the robot, plan the motion trajectory of the patient's affected leg. Specific steps are as follows:

[0078] Step 2.1: Establish a motion trajectory model based on the lower limb rehabilitation robot trajectory data obtained in Step 1. The mathematical formula for the trajectory model is:

[0079]

[0080] Where, x, It refers to the robot's trajectory, velocity, and acceleration, α. x and β x It is a proportionality constant and β x =α x / 4,τ s The duration of motion can be represented by τ. s The convergence speed of the trajectory is adjusted, where g is the target position of the trajectory, and g is used to change the amplitude and frequency of the trajectory. f(x) is used to adjust the endpoint and shape of the trajectory, specifically:

[0081]

[0082] in It is the Gaussian function, N represents the number of Gaussian functions, ω i These are the weights of each basis function that determine the shape of the gait trajectory.

[0083] Step 2.2: Add a human-computer interaction term to the trajectory model in Step 2.1, and describe the human-computer interaction using an admittance model. Specifically:

[0084]

[0085] Where F is the interaction force between humans and machines, and k h and d h These are impedance model parameters, x and It is the displacement and acceleration output by the admittance model based on the interaction force between the human and the machine.

[0086] Step 2.3: Let x in step 2.2 d=0, adjust the output displacement x of the admittance model according to the human-machine interaction force F, and let the target position g of the trajectory model in step 2.1 be x, so as to adjust the target position of the trajectory model according to the human-machine interaction force and plan the movement trajectory of the patient's affected leg.

[0087] In this embodiment, the mirror-based rehabilitation strategy structure is as follows: Figure 3 As shown, the mirror rehabilitation strategy structure consists of three parts: the healthy leg part, the control system, and the affected leg part. The control system includes a perception layer, a conversion layer, and an execution layer. Perception layer: identifies the human movement intention; Conversion layer: plans the movement trajectory based on the established trajectory model; Execution layer: transmits the planned movement trajectory to the affected leg part and performs compliant control on the affected leg. The specific process is as follows: (1) The patient's healthy leg interacts with the robot, and the interaction force is input into the control system. (2) Based on the interaction force input into the control system, the perception layer identifies the patient's movement intention and determines the target point of the movement trajectory based on the magnitude of the interaction force, and transmits the obtained target point to the conversion layer of the control system. (3) The conversion layer inputs the target point transmitted from the perception layer into the trajectory model and plans the movement trajectory of the healthy leg based on the target point. (4) The role of the execution layer is to transmit the movement trajectory of the healthy leg to the affected leg. Finally, the affected leg is compliantly controlled based on the movement trajectory output by the execution layer.

[0088] Step 3: Design a trajectory tracking controller to accurately track the movement trajectory of the affected leg in Step 2.

[0089] Step 3.1: Establish the dynamic model of the lower limb rehabilitation robot, specifically as follows:

[0090]

[0091] Where, θ, and These are the hip joint angle, angular velocity, and angular acceleration of the rehabilitation robot, respectively. Furthermore, M(θ) is the inertia matrix. Here, G(θ) is the Coriolis and centrifugal force matrix, G(θ) is the robot's gravity matrix, and τ is the control torque applied to the joints. hr It is the torque for human-computer interaction.

[0092] Let x1 = θ, u = τ, the dynamic equations of the lower limb rehabilitation robot are rewritten as state-space equations:

[0093]

[0094] in,

[0095] f(x1,x2)=-M -1 (x1)(C(x1,x2)+G(x1))

[0096] M(x1)=(m1+m2)l 2 / 3

[0097] C(x1,x2)=-2m2l 2 sinx1cosx1

[0098] G(x1)=((m1+m2)glcosx1) / 2

[0099] m1 and m2 are the masses of the thigh and lower leg connecting rod, and l is the length of the thigh and lower leg connecting rod.

[0100] Step 3.2: Design the trajectory tracking controller, specifically as follows:

[0101] Let z1 = x1 - x 1d Differentiate with respect to z1:

[0102] Using Lyapunov functions V1 represents the constructed Lyapunov function, differentiated with respect to time:

[0103]

[0104] make have to therefore,

[0105]

[0106] Let z² = x² - x 2d Differentiate it:

[0107] Define Lyapunov functions V2 represents the function constructed in Lyapunov form, differentiated with respect to time:

[0108]

[0109] make It is τ hr The estimated value.

[0110] Define Lyapunov functions V3 represents a function constructed in Lyapunov form, differentiated with respect to time:

[0111]

[0112] make Obtain the estimated value of human-computer interaction force

[0113] The final trajectory tracking controller is:

[0114]

[0115] In this embodiment, the structural parameters of the lower limb rehabilitation robot are as follows: m1 = 2.46 kg, m2 = 3.14 kg, l1 = 0.51 kg, l2 = 0.56 kg.

[0116] Step 4: Design an adaptive admittance controller to correct the input trajectory of the trajectory tracking controller in Step 3, thereby improving the robot's motion compliance. The specific steps are as follows:

[0117] Step 4.1: Establish the admittance model, specifically as follows:

[0118]

[0119] Where ΔT is the change in the interaction torque between the human and the robot, M, B, and K are the inertia, damping, and stiffness of the admittance model, respectively, and θ, and These are the robot's angular displacement, angular velocity, and angular acceleration.

[0120] Step 4.2: Design the adaptive law for admittance parameters, specifically as follows:

[0121]

[0122]

[0123] Where B0 and K0 are base values, a j and b j (j=1,2,3) are the coefficients for adjusting the damping and stiffness of the robot admittance model, respectively, and B lj and K lj (j = 1, 2) represent the upper and lower thresholds for damping and stiffness. T is the interaction torque between the human body and the robot, setting thresholds for the damping and stiffness parameters to ensure the stability of the control system. As the robot's angular displacement and interaction torque increase, the stiffness and damping decrease. This ensures that the lower limb rehabilitation robot can be safely driven when the patient is in an uncomfortable position, thus preventing secondary injury to the patient.

[0124] Step 5: Combine the trajectory tracking controller from Step 3 with the adaptive admittance controller from Step 4 to design a hierarchical control strategy and define the human-machine interaction force threshold. Different control strategies are selected based on different interaction forces. If the force is within the threshold range, the controller from Step 3 is used to complete the training task; otherwise, the controller from Step 4 is activated to correct the input trajectory of the controller from Step 3, achieving compliant control of the patient's affected leg. The specific steps are as follows:

[0125] Step 5.1: Design a hierarchical control strategy, including an upper-level adaptive admittance controller and a lower-level position tracking controller. When the human actively interacts with the robot, the upper-level admittance controller adjusts the motion trajectory of the lower limb rehabilitation robot and transmits the adjustment amount as input to the lower-level position tracking controller to control the robot to drive the patient's affected leg to track the trajectory.

[0126] Step 5.2: Based on the interaction between humans and robots, define the range of plantar pressure. According to the different magnitudes of the interaction forces between humans and robots, divide the interaction between patients and robots into three regions: passive movement range, cooperative movement range, and abnormal movement range. When the human-machine interaction force is within the passive movement range, the robot performs trajectory tracking tasks; when the human-machine interaction force is within the cooperative movement range, the patient can actively apply force to change the robot's movement state; when the human-machine interaction force is within the abnormal movement range, the robot stops moving.

[0127] In this embodiment, in order to more accurately execute different rehabilitation training modes, the range of human-computer interaction force is divided into different regions, and the division method is as follows: Figure 4 As shown, in Figure 4 In the diagram, the pressure curve represents the contact force between the patient and the robot during passive rehabilitation training, when the patient does not exert any active force on the robot. Based on the force exerted by the patient on the rehabilitation robot during the training process, the pressure curve is divided into three distinct regions. d P(t) represents the interaction force between the patient and the rehabilitation robot during passive rehabilitation training. P(t) is the active force exerted by the patient on the rehabilitation robot during rehabilitation training. d(t) is the interaction force between P and the robot. d The distance between d(t) and P(t). d(t)∈[0,r H Defined as the passive range of motion, the lower limb rehabilitation robot uses a lower-level position controller to perform passive rehabilitation training for patients. d(t)∈(r) H ,r R [ ] represents the range of cooperative motion; when P(t) exceeds a predetermined threshold r H At that time, the upper admittance controller will be activated to adjust the motion trajectory, (P(t)-r H d(t) is the active force applied to the robot to adjust its trajectory until P(t) returns to the passive training range. H (+∞) indicates an abnormal range of motion, signifying abnormal movements by the patient during training. The rehabilitation robot will then stop the ongoing rehabilitation training, which is crucial for the patient's safety during the training process.

[0128] With simple modifications, the hardware requirements for a mirror rehabilitation strategy can be met by adapting a lower limb rehabilitation robot. Mirror rehabilitation can effectively improve the coordination of the patient's legs; compliant control of the affected leg can effectively prevent secondary injury during training, improving the safety and comfort of the patient's rehabilitation training. The same invention applies to upper limb rehabilitation robots made with the same modifications.

[0129] A mirror-compliant control device for a lower limb rehabilitation robot based on dynamic motion primitives, including

[0130] The filtering module is used to collect the robot end-effector trajectory data on the same side as the patient's healthy leg during flexion and extension training, filter the data, and save it to the computer.

[0131] The trajectory generation module is used to establish a motion trajectory model of the lower limb rehabilitation robot based on the filtered data, using the dynamic motion primitive method combined with the admittance model, and to plan the motion trajectory of the patient's affected leg based on the interaction force between the healthy leg and the robot.

[0132] The policy control module includes:

[0133] A trajectory tracking controller is used to accurately track the movement trajectory of the affected leg;

[0134] An adaptive admittance controller is used to correct the input trajectory of the trajectory tracking controller;

[0135] By combining a trajectory tracking controller with an adaptive admittance controller, a hierarchical control strategy is designed, and a human-machine interaction force threshold is defined. Different control strategies are selected based on different interaction forces.

[0136] Furthermore, the trajectory generation module is used to establish a motion trajectory model based on the robot's end-effector trajectory data;

[0137] Add a human-computer interaction term to the motion trajectory model and use an admittance model to describe the human-computer interaction.

[0138] Set initial parameters and target position of trajectory model, adjust target position of trajectory model according to human-computer interaction force, and plan the movement trajectory of patient's affected leg.

[0139] An electronic device includes a memory in which processors are interconnected, the memory storing computer instructions, and the processors executing the computer instructions to perform the mirror compliance control method for a lower limb rehabilitation robot based on dynamic motion primitives as described above.

[0140] An electronic device includes a memory in which processors are interconnected. The memory stores computer instructions, and the processors execute the aforementioned mirror compliance control method for a lower limb rehabilitation robot based on dynamic motion primitives by executing the computer instructions.

[0141] A computer storage medium storing computer instructions, the computer instructions being used to cause a computer to execute the above-described mirror compliant control method for a lower limb rehabilitation robot based on dynamic motion primitives.

[0142] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0143] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0144] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0145] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0146] The above-described embodiments provide a detailed explanation of the technical solution of the present invention. It is worth noting that these embodiments are merely specific examples of the present invention and are not intended to limit the present invention. Any modifications and improvements made within the scope of the present invention should be included within the protection scope of the present invention.

Claims

1. A mirror-compliant control system for a lower limb rehabilitation robot based on dynamic motion primitives, characterized in that, include: The filtering module is used to collect the robot end-effector trajectory data on the user's first leg during flexion and extension training, filter the data, and save it to the computer. The trajectory generation module is used to establish a motion trajectory model of the lower limb rehabilitation robot based on the filtered data, using the dynamic motion primitive method combined with the admittance model. Based on the interaction force between the first leg and the robot, the motion trajectory of the user's second leg is planned. The policy control module includes: A trajectory tracking controller is used to accurately track the movement trajectory of the user's second leg; An adaptive admittance controller is used to correct the input trajectory of the trajectory tracking controller; By combining a trajectory tracking controller with an adaptive admittance controller, a hierarchical control strategy is designed, and a human-machine interaction force threshold is defined. Different control strategies are selected according to different interaction forces. If the force is within the threshold range, the trajectory tracking controller is used to complete the training task. If the force is outside the threshold range, the adaptive admittance controller is activated to correct the input trajectory of the trajectory tracking controller, thereby achieving compliant control of the user's second leg.

2. The mirror-compliant control system for a lower limb rehabilitation robot based on dynamic motion primitives according to claim 1, characterized in that, The trajectory generation module includes: The input module is used to build a motion trajectory model based on the robot's end effector trajectory data; The computation module is used to add human-computer interaction terms to the motion trajectory model and describe the human-computer interaction using an admittance model. The planning module sets initial parameters and the target position of the trajectory model, adjusts the target position of the trajectory model based on human-computer interaction force, and plans the movement trajectory of the user's second leg.

3. The mirror-compliant control system for a lower limb rehabilitation robot based on dynamic motion primitives according to claim 1, characterized in that: A motion trajectory model is established based on the obtained robot end-effector trajectory data. The mathematical formula for the motion trajectory model is as follows: in, , , It refers to the robot's trajectory, speed, and acceleration. and It is a proportionality coefficient and ; Indicates the duration of motion, by Adjust the convergence speed of the trajectory; It is the target location of the trajectory, through To change the amplitude and frequency of the trajectory; Used to adjust the endpoint and shape of the trajectory, specifically: in It is the Gaussian function. Indicates the number of Gaussian functions. These are the weights of each basis function that determine the shape of the gait trajectory; A human-computer interaction term is added to the motion trajectory model, and the human-computer interaction is described using an admittance model, wherein the admittance model is: in For human-computer interaction, and These are impedance model parameters. and It is the displacement and acceleration output by the admittance model based on the interaction force between the human and the machine; Let the reference trajectory of the admittance model Based on the interaction force between humans and machines Adjusting the output displacement of the admittance model And set the target position of the trajectory model This allows the trajectory model to be adjusted based on human-computer interaction forces to plan the movement trajectory of the user's second leg.

4. The mirror-compliant control system for a lower limb rehabilitation robot based on dynamic motion primitives according to claim 1, characterized in that: A dynamic model of the lower limb rehabilitation robot is established, represented as follows: in, , and These are the hip joint angle, angular velocity, and angular acceleration of the rehabilitation robot; It is the inertia matrix. It is the Coriolis and centrifugal force matrix. It is the robot's gravity matrix. It is the control torque applied to the joint. It is the torque for human-computer interaction; make , , The dynamic equations of the lower limb rehabilitation robot are rewritten as state-space equations: in, and It is the mass of the connecting rod between the thigh and lower leg. It is the length of the connecting rod between the thigh and lower leg; Design a trajectory tracking controller and compensate for the human-computer interaction force. The estimated value of the human-computer interaction force is: ， The trajectory tracking controller is: 。 5. The mirror-compliant control system for a lower limb rehabilitation robot based on dynamic motion primitives according to claim 1, characterized in that: Establish the admittance model, specifically as follows: in, It is the change in the torque of the interaction between humans and robots. , and These are the inertia, damping, and stiffness of the admittance model, respectively. , and These are the robot's angular displacement, angular velocity, and angular acceleration, respectively. Design an adaptive law for admittance parameters, specifically: in and It is the base value. and These are the coefficients for adjusting the damping and stiffness of the robot's admittance model, respectively. and These are the upper and lower thresholds for damping and stiffness. It is the torque of interaction between the human body and the robot.

6. The mirror-compliant control system for a lower limb rehabilitation robot based on dynamic motion primitives according to claim 1, characterized in that: A hierarchical control strategy is designed, including an upper-level adaptive admittance controller and a lower-level position tracking controller. When the user interacts with the robot, the upper-level admittance controller adjusts the motion trajectory of the lower limb rehabilitation robot and transmits the adjustment amount as input to the lower-level position tracking controller, which controls the robot to drive the user's second leg to track the trajectory. Based on the interaction between the user and the robot, the range of plantar pressure is defined. According to the different magnitudes of the interaction force between the user and the robot, the interaction between the user and the robot is divided into regions, and corresponding control strategies are executed according to the corresponding regions.

Images