CN101301250A - Control strategy for interactive rehabilitation training of five-degree-of-freedom exoskeleton upper limb rehabilitation robot - Google Patents

Abstract

The invention provides a five-degree of freedom exoskeleton-type upper limb recovery robot interactive recovery training control strategy. The strategy comprises two training patterns corresponding to the different recovery period of a patient, i.e. a passive interactive recovery training control strategy and an active-auxiliary interactive recovery training control strategy; during passive movement, a surface myoelectric signal of relevant muscles on a healthy upper limb of a patient is picked up and taken as the movement intention of the patient, thereby controlling a robot to drive a diseased side to realize movement passive training; during active movement, the movement intention of an upper limb of a human body is judged through acquiring in real time the arthrosis force moment generated by a diseased limb acting on the robot during movement, and an applied force is converted into the velocity quantity of the tail end of a mechanical arm by means of a proportional controller, thereby driving the robot to follow the intention so as to carry out recovery active-auxiliary training of the diseased limb. The five-degree of freedom exoskeleton-type upper limb recovery robot interactive recovery training control strategy can provide all-around recovery training movement for a clinic hemiplegia patient, thereby increasing the activeness of the patient in recovery training and the confidence of the patient in recovery; meanwhile, the recovery training control strategy also increases the attractiveness of a recovery process and promotes recovery efficacy.

Description

Control strategy for interactive rehabilitation training of five-degree-of-freedom exoskeleton upper limb rehabilitation robot

(1) Technical field

The invention relates to a rehabilitation robot technology, in particular to a five-degree-of-freedom exoskeleton type upper limb rehabilitation robot interactive rehabilitation training control technology.

(2) Background technology

At present, the traditional upper limb clinical rehabilitation methods are mostly as shown in Figure 1. Rehabilitation physicians guide patients to complete continuous passive movements one-on-one with bare hands or with assistive devices, or give appropriate resistance or assistance to guide patients to move. The increase in the incidence of stroke has led to more and more patients with hemiplegia. The heavy and high-intensity training tasks have brought a heavy burden to the treating physicians. Improvements in treatment options. On the other hand, factors such as the therapist's subjective consciousness, physical strength, and mood play a leading role in the training process, ignoring the patient's own movement intention, coupled with the impairment of the language and cognitive functions of the patient after the onset of the disease, weakening the Interaction between doctors and patients for rehabilitation treatment process.

Therefore, combining robots and related technologies with clinical rehabilitation medicine, a rehabilitation robot is designed to replace physicians to complete rehabilitation training for hemiplegic patients. Although the existing upper limb rehabilitation robots can reduce the burden on doctors, improve efficiency, and provide objective and effective data basis for the formulation and optimization of rehabilitation programs for patients in different periods, there are still the following problems: the training actions provided for patients with affected limbs Most of them are simple straight or curved trajectories, with a small range of motion and a single type; they often apply predetermined programs or apply predetermined resistance/assist to achieve passive movement or auxiliary movement of the upper limbs, which also ignores the movement intention of the patient's limbs, which is not conducive to stimulation The patient's active exercise improves the patient's interest in exercise and builds the patient's exercise confidence.

(3) Contents of the invention

The purpose of the present invention is to provide a kind of rehabilitation training movement that can provide a full range of clinical hemiplegia patients, thereby improving the initiative of the patient's rehabilitation training, making it strengthen the recovery confidence, and also enhancing the attractiveness of the recovery process and promoting A five-degree-of-freedom exoskeleton upper limb rehabilitation robot interactive rehabilitation training control strategy for rehabilitation efficacy.

The purpose of the present invention is achieved in this way: two training modes corresponding to different recovery periods of the patient: passive and active-assisted interactive rehabilitation training control strategies, in the passive movement, extract the surface myoelectricity of the relevant muscles of the healthy side of the patient's upper limbs The signal serves as the patient's movement intention, and the robot is controlled to drive the affected side to achieve passive training of movement; in the active movement, the movement intention of the upper limb of the human body is judged by obtaining the joint torque generated by the affected limb acting on the robot in real time during the movement, and the proportional control is used The device converts the applied force into the velocity of the end of the mechanical arm, and then drives the robot to follow the intention to perform active-assisted training for the rehabilitation of the affected limb.

The present invention also has some technical characteristics:

1. In the early and middle stages of the recovery of the patient, including the flaccid phase and the spastic phase, passive exercise training is applied, and according to the characteristics of the upper limb damage of the hemiplegic patient, the surface electromyography signals of the four muscles of the uninjured upper limb of the hemiplegic patient are extracted ( sEMG) is used as the control signal of the rehabilitation robot, including the middle deltoid, anterior deltoid, biceps brachii and brachioradialis corresponding to each surface electromyographic electrode in turn, using the AR parameter model and BP neural network to understand the patient's movement intention, Drive the mechanical arm to assist the upper limb of the affected side to realize the undetermined rehabilitation training action;

2. In the middle and late stages of the patient’s recovery, including the spasticity period and the improvement period, the active movement training is applied, which is a typical double-loop control system structure, the outer loop is the force control loop at the end of the mechanical arm, and the inner loop is the joint speed The control loop, at the sampling time t, uses the joint torque sensor to collect the torque value exerted on each joint by the active movement of the hemiplegic patient, and at the same time calculates the robot no-load torque value as a function of the joint angle at the current moment; send the two into The robot end force synthesizer obtains the end force, that is, the estimated value of the force exerted by the hemiplegic patient on the end point of the robot, so as to obtain the patient's motion intention; use the proportional controller to convert it into the robot end motion speed, and pass the kinematics Jacobian The inverse matrix is mapped to the angular velocity command of each joint of the robot; the angular velocity command is input into the inner loop controller of the robot joint, and the servo motor of each joint is controlled to achieve a stable angular velocity output, thereby driving the robot to follow the patient's active movement intention and realize rehabilitation training force-assisted control.

The interactive rehabilitation training control strategy of the five-degree-of-freedom exoskeleton type upper limb rehabilitation robot of the present invention adopts passive and active-assisted training modes for different recovery situations of clinical hemiplegia patients, and obtains the surface muscles of relevant muscles during the patient's exercise process. The electric signal and the torque information generated during the movement process can predict the patient's movement intention, and then drive the robot to assist the patient's affected limb to realize the training of various single, multi-joint and daily life actions. The control strategy introduces the patient's active movement intention, which has a good effect on stimulating the patient's active movement, exciting the diseased brain, and promoting the efficacy of rehabilitation treatment.

The five-degree-of-freedom exoskeleton upper limb rehabilitation robot system consists of a robotic arm body, a drive motor, a torque sensor, a surface electromyography electrode, a control cabinet, a mounting frame, and a seat. The computer in the control cabinet drives the motors installed on each rotation axis. The motor is driven so that the main body of the robotic arm drives the patient's affected limb to perform rehabilitation training exercises. The torque sensors at each joint measure the torque information during the exercise process, and the surface electromyography electrodes measure the surface electromechanical signals of the muscles related to the exercise. The structural form of the exoskeleton bilateral structure reduces the mass while ensuring the rigidity. At the same time, it provides single-joint and multi-joint compound rehabilitation training exercises for hemiplegic patients. To provide hemiplegia patients with simple daily life functional actions - training of eating and trousers lifting, the invention introduces the man-machine interactive hemiplegia rehabilitation training control strategy of the patient's own movement intention. This rehabilitation robot provides a full range of rehabilitation training exercises for patients with clinical hemiplegia. From the perspective of training actions, it includes the training of all joints of the upper limbs, including passive movements from large shoulder joints to small wrist joints and active-assist Exercise, supplemented by simple daily life actions - training of eating and lifting pants, deepen and apply each single joint action training. From the point of view of the training mode, the rehabilitation robot adopts two training modes corresponding to different recovery periods of the patients in each exercise provided: passive and active-assisted interactive rehabilitation training control strategies. That is, in the process of passive and active-assisted training, the movement intention of the patient's affected limb is reflected by the surface electromyographic signals of the relevant muscles of the upper limb and the joint torques generated, and the robot is controlled to assist the patient's limb in rehabilitation training according to the intention. . Thereby, the initiative of the patient's rehabilitation training is improved, the confidence of treatment is enhanced, the attractiveness of the treatment process is also enhanced, and the efficacy of rehabilitation treatment is promoted.

(4) Description of drawings

Fig. 1 is the schematic diagram of traditional clinical rehabilitation method;

Fig. 2 is a schematic diagram of the upper limb rehabilitation robot system;

Fig. 3 is the schematic diagram of electric control cabinet of rehabilitation robot;

Fig. 4 is a structural diagram of the upper limb rehabilitation mechanical arm;

Fig. 5 is a schematic diagram of surface electromyography electrode distribution;

Fig. 6 is a passive and active two-stage interactive rehabilitation training control strategy;

Fig. 7 is a block diagram of the control principle of the main power assisted rehabilitation training.

(5) Specific implementation methods

The present invention will be further described below in conjunction with accompanying drawing and specific embodiment:

With reference to FIG. 2 , the system rehabilitation training device of this embodiment is composed of a rehabilitation robot arm body 1 , a drive motor, an adjustable lifting mounting frame 2 , a seat 3 and a control cabinet (combined with FIG. 3 ). Combined with Figure 4, the body of the rehabilitation robotic arm is designed as a duralumin exoskeleton bilateral structure, with shoulder abduction/adduction I, shoulder flexion/extension II, elbow flexion/extension III, wrist flexion/extension IV and For the 5 degrees of freedom of wrist internal/external rotation V, 5 drive motors are respectively installed on the axis of each rotational degree of freedom to drive the activities of each degree of freedom. From the perspective of clinical rehabilitation, the rehabilitation training device provides patients with single-joint movement of each joint and multi-joint compound movement in three-dimensional space, and provides simple and basic daily life action training: eating, lifting pants. Four torque sensors (two on the shoulder, one on the elbow, and one on the flexion and extension of the wrist) installed on the shoulder, elbow, and wrist cascaded with the drive motor respectively measure the torque information applied by the human body to the robot during the movement process. The surface electromyographic electrodes 4, 5, 6, and 7 (combined with Figure 5) are used to obtain surface electromyographic signals that directly reflect muscle movement during exercise. Both signals are used to monitor the movement of the robot and are used in different modes of interactive rehabilitation control strategy.

Combined with Figure 6, in each movement provided by the robot, two training modes corresponding to different recovery periods of the patient are adopted: passive and active-assisted interactive rehabilitation training control strategies. In the early and middle stages of the patient's recovery (including the flaccid and spastic stages), passive exercise training was applied, and according to the characteristics of the upper limb damage of the hemiplegic patient, the four muscles of the healthy side of the hemiplegic patient's upper limb were extracted (combined with Figure 5, each surface EMG The surface electromyographic signal (sEMG) of the middle deltoid muscle, front deltoid muscle, biceps brachii, and brachioradialis corresponding to the electrodes in turn is used as the control signal of the rehabilitation robot, and the AR parameter model and BP neural network are used to understand the patient's movement intention , to drive the mechanical arm to assist the upper limb of the affected side to realize the undetermined rehabilitation training action. In the middle and later stages of the patient's recovery (including the spasticity period and the improvement period), active exercise training is applied, such as the control principle block diagram of active force-assisted rehabilitation training (combined with Figure 7). This is a typical double-loop control system structure, and the outer loop is mechanical. The force control ring at the end of the arm, and the inner ring is the joint speed control ring. The specific working principle is: at the sampling time t, use the joint torque sensor to collect the torque value exerted on each joint by the active movement of the hemiplegic patient, and at the same time calculate the robot no-load torque value as a function of the joint angle at the current moment; combine the two Send it to the end force synthesizer of the robot to obtain the end force, that is, the estimated value of the force exerted by the hemiplegic patient on the end point of the robot, so as to obtain the patient's movement intention; use the proportional controller to convert it into the end movement speed of the robot, and through kinematics The Jacobian inverse matrix is mapped to the angular velocity command of each joint of the robot; the angular velocity command is input into the inner loop controller of the robot joint, and the servo motor of each joint is controlled to achieve a stable angular velocity output, thereby driving the robot to follow and realize the patient's active motion intention Force-assisted control for rehabilitation training.

Claims (3)

1, a kind of five-freedom degree dermaskeleton type upper limb rehabilitation robot interactive rehabilitation training control policy, it is characterized in that: the different convalescent two kinds of training modes of corresponding patient: the interactive rehabilitation training control policy of passive type and active one assist type, in passive exercise, the surface electromyogram signal that extracts patient's healthy side upper limb related muscles is intended to as patient moving, and the control robot drives the passive exercise that Ipsilateral is realized motion; In the active exercise, suffer from limb and act on the motion intention that joint moment that robot produces is judged human upper limb by obtaining in real time in the motor process, and the proportion of utilization controller will apply the speed amount that power is converted into the mechanical arm end, and then the driven machine people follows intention and carries out active--the supplemental training of rehabilitation to suffering from limb.

2, five-freedom degree dermaskeleton type upper limb rehabilitation robot interactive rehabilitation training control policy according to claim 1, it is characterized in that: described morning of recovering the patient, mid-term, comprise the phase of collapsing from physical exhaustion and spasm period, use continuous passive motion, the characteristics impaired according to hemiplegic patient by upper, extract the control signal of four muscle surface electromyographic signals of healthy side upper limb (sEMG) of hemiplegic patient as healing robot, comprise in the corresponding successively triangular muscle of each surface myoelectric electrode, before the triangular muscle, biceps brachii m. and brachioradialis, utilize AR parameter model and BP neutral net to understand patient's motion intention, the auxiliary Ipsilateral upper limb of driving device arm is realized rehabilitation training action undetermined.

3, five-freedom degree dermaskeleton type upper limb rehabilitation robot interactive rehabilitation training control policy according to claim 1 and 2, it is characterized in that: described in patient's recovery, later stage, comprise spasm period and improvement phase, application active exercise training, it is a typical dicyclo control system structure, outer shroud is the terminal power control loop of mechanical arm, internal ring is the joint velocity control loop, at sampling instant t, utilize the joint moment sensor collection to be applied to moment values on each joint by hemiplegic patient's active exercise, calculate unloaded moment values simultaneously as the robot of each joint angles function of current time; The two is sent into robot end's power synthesizer obtain terminal power, promptly the hemiplegic patient is applied to the estimated value of the power at robot end's point place, thereby obtains patient's motion intention; The proportion of utilization controller is translated into robot end's movement velocity, and is mapped as the angular velocity instruction in each joint of robot by kinesiology Jacobi inverse matrix; Angular velocity instruction is input in the joint of robot in the middle of the ring controller, control the servomotor in each joint and realize stable angular velocity output, thereby the driven machine people follows and realize the power assist control of rehabilitation training to patient's active exercise intention.

Images