Hand rehabilitation exoskeleton robot

Abstract

The invention designs a hand rehabilitation exoskeleton robot, and belongs to the field of robots. Five fingers can be recovered at the same time. Each finger comprises a wearing module worn on each knuckle, an SEA series elastic driving module, a multi-connecting-rod coupling module and a sensor module. According to the wearable module, all knuckles of a person serve as rigid connecting rods to be connected with the exoskeleton, and the wearable module can adapt to the human knuckles of different sizes by adjusting sliding blocks. The SEA series elastic driving module not only can realize bidirectional flexible driving, but also can realize force feedback without a force sensor in combination with an angle sensor, so that the cost and the size are saved; the multi-connecting-rod coupling module drives three joints of a single finger in a single-degree-of-freedom mode through the coupling relation between connecting rod mechanisms, and the risk that the joints are injured is reduced. The sensor module comprises an angle sensor, a myoelectricity sensor and an electroencephalogram sensor, the angle sensor measures real-time joint angle changes, the myoelectricity sensor collects myoelectricity signals, and full-drive composite impedance control of electroencephalogram and myoelectricity fusion is achieved in combination with concentration degree information extracted by the electroencephalogram sensor.

Description

Technical Field

[0001] The present invention designs a hand rehabilitation exoskeleton robot, which specifically assists patients with hand movement disorders to exercise and achieve full-cycle rehabilitation, and belongs to the field of robots. Background Art

[0002] The human hand is the main effector organ for humans to carry out various types of movements and plays an indispensable role in human daily life. Studies have shown that compared with other primates, human hands have higher flexibility. Hands are crucial in people's daily activities, and some sudden diseases or accidental injuries lead to different degrees of loss of human hand function, such as rheumatoid arthritis and stroke (commonly known as stroke), which not only brings inconvenience to patients' lives and work, but also causes physical and psychological harm to patients. Fortunately, the human motor system can restore the damaged function through high-intensity and repetitive rehabilitation training. In order to allow stroke patients to better return to society and family, systematic rehabilitation training is particularly important and will directly determine the possible degree of recovery of patients. The hand exoskeleton can assist the elderly and patients with motor dysfunction to grasp daily objects, and can also perform rehabilitation training to prevent muscle atrophy and restore the movement ability of the affected limb. This exoskeleton is an intelligent rehabilitation robot that can be connected in parallel to the human hand through a rigid structure, relying on reasonable mechanism design and control algorithm to provide patients with a more efficient, safe and reliable rehabilitation experience.

[0003] Traditional hand exoskeletons still have many shortcomings, such as the inability to ensure instantaneous alignment of joint centers, which can easily cause harm to patients and poor physical sensation; the system transmission is not precise enough; it cannot adapt to the sizes of different people, and there is no structural optimization process. Most of them lack sensors for biological signals, and have not achieved better human-machine interaction and compatibility between the human body and the robot.

[0004] In the invention patent number CN103315880B, a hand exoskeleton rehabilitation system based on memory alloy drive is disclosed. The present invention uses five shape memory alloy springs as the driving device to independently drive the movement of five fingers, and adopts a four-bar linkage as the transmission mechanism of the exoskeleton. It is required that the degrees of freedom of the exoskeleton and the degrees of freedom of each joint of the human hand are highly matched, otherwise it is easy to cause safety accidents, poor physical sensation, and secondary injuries.

[0005] In the invention patent number CN109276408A, an upper limb hand exoskeleton rehabilitation robot is disclosed. The device allows the patient to watch virtual reality images, obtains the user's movement intention based on EEG, and then uses the exoskeleton to drive the human hand to complete the virtual reality task. The exoskeleton glove is simple to set up, without integrating sensors to measure the human-machine contact force and joint angle of the hand; five servos are used to wind the wire rope to pull the finger movements of the five fingers, and the control process is relatively cumbersome; in addition, there is a large delay in obtaining the user's movement intention through EEG, and the real-time control is not good.

[0006] In the invention patent with patent number CN209092054U, a hand exoskeleton capable of autonomous rehabilitation training based on wire drive and mirror synchronous simulation is disclosed. The system consists of an active hand exoskeleton and a mirror hand exoskeleton, and the overall structure consists of a cartilage-type elastic material and a small motor group, a control circuit, a measurement circuit, a bending deformation sensor, and a wireless module. A motor group consisting of five motors is used on the palm side to pull the rope to bend the hand; five springs are used on the back of the hand to passively pull the rope to stretch the hand; the control process is relatively complicated; secondly, the rope is simply wound on the hand, which easily leads to uneven force distribution; finally, the sensor information of the mirror hand is used to obtain the user's movement intention, which makes the process complicated, and the entire operation process can be completed on the same arm.

[0007] Therefore, in view of the problems existing in the existing hand exoskeletons, it is necessary to design a system that can achieve individual differentiated adaptation in structure, with a compact and comfortable mechanical structure; realize biomechanical multi-source fusion in perception, and force feedback without force sensors; and a system with brain and myoelectric fusion control to enhance the motor function of the human hand and accelerate the recovery of hand function. Summary of the invention

[0008] The purpose of the present invention is to provide an under-actuated, lightweight and soft hand rehabilitation exoskeleton robot that realizes individual differentiation adaptation, biomechanical and electrical multi-source fusion, force feedback without force sensors, and impedance control.

[0009] The present invention achieves the above-mentioned purpose through the following technical solutions: a hand exoskeleton robot for assisting the rehabilitation of stroke patients, including an under-actuated multi-link coupling mechanism. In order to improve the user's body sensation, it is based on rigid links and is well adapted to the sizes of different human hands. The overall structure of the exoskeleton is designed based on the structural characteristics of the human hand and the laws of joint movement. The MCP (mecarpophalangeal joint) and PIP (proximal interphalangeal joint) of the human hand do not have a definite coupling relationship. The angle relationship between the MCP and DIP (distal interphalangeal joint) of different people's fingers is different, but the DIP and PIP joints maintain a certain working range. , so if you want to use a push rod to achieve under-actuation, you have to let the connecting rod perform coupled motion. The first joint is driven by a predetermined trajectory. Due to the angle bisector theorem, when moving on a circular arc trajectory, by adding a slider degree of freedom, the human hand joint can be kept consistent with the instantaneous center of the exoskeleton to maintain the comfort of the human hand. At the same time, the extra degree of freedom can be adjusted by adjusting the position of the slider in the initial position, so that the exoskeleton can adapt to the differences of different human fingers while keeping the instantaneous center consistent. The second and third joints are connected by a multi-link coupling mechanism. The different knuckles of human fingers are regarded as the links in the multi-link. The multi-link coupling mechanism is composed of a mid-end four-link and a distal four-link, which determine the angles of the MCP and PIP joints respectively. The curved link with three rotation joints and a fixed angle belongs to both the mid-end four-link mechanism and the distal four-link mechanism. As the motor stroke length increases, the position of the mid-end four-link mechanism changes, and the movement of the mid-end four-link mechanism affects the position of the distal four-link mechanism. Due to the combination of two four-bar linkages and a bent link, the movement of the motor stroke controls the angles of the MCP and PIP joints at the same time. Since it is a multi-link mechanism and the finger knuckles serve as links, the joint center also serves as a rotation pair of the linkage mechanism, which can keep the instantaneous center of motion of the exoskeleton and the human hand joint consistent. The initial position of the exoskeleton can be adjusted through self-locking nuts and spring guides to adapt to the size of different human fingers.

[0010] Furthermore, the designed exoskeleton is structurally optimized to better meet the natural flexion and extension process of human fingers. The angle coupling of the PIP and DIP joints is optimized through mathematical modeling to determine the optimized size. At the same time, for the proximal knuckles, since the connecting rods are prone to interference during work, the structure of the proximal connecting rods is optimized for the purpose of non-interference, so that there will be no interference in the complete working space. According to the force of the fingers in the rehabilitation stage, the appropriate stiffness of the compression spring is selected, and the SEA series elastic driver is designed to achieve bidirectional compliant drive of the extension and bending movements of the exoskeleton. Each finger is driven by a linear push rod, which is connected to the SEA series elastic driver and connected in series through two compression springs to achieve bidirectional compliant drive.

[0011] Furthermore, a multi-sensor fusion perception system is adopted, which is composed of fingertip end sensors, rotary potentiometers, myoelectric bracelets, EEG sensors and force sensing based on SEA series elastic drive sensors. Angle sensors are used at each joint to track and record the joint movement trajectory and read the angle changes during joint movement. Since the length of the connecting rod of the connecting rod mechanism will change after adjusting the wearable parts, the angle of the initial position is also different. The angle of the initial position is obtained through kinematic inverse solution, and the working range of the push rod is adjusted in real time according to the differences of different people. The spring of the SEA series elastic drive is calibrated for stiffness, and the output end of the direct-acting push rod is equipped with a spring compression amount through the SEA series elastic drive to collect the pulling pressure, so as to realize the force feedback of the forceless sensor, reduce the number of sensors, and reduce the overall size and mass of the exoskeleton.

[0012] Furthermore, a force feedback based on a forceless sensor with a SEA series elastic actuator is adopted, where a and b are constants determined by the mechanical structure, θ is measured by an angle potentiometer, and l 1 Measured by the electric push rod encoder, according to the geometric relationship:

[0013]

[0014] The initial length of the spring is known, so the spring deformation can be calculated, and then the spring force can be calculated. The spring force reflects the magnitude of the grasping force. Fig.16 As shown, the electromyographic signal is collected, the electromyographic signal is subjected to a preamplifier circuit, low-pass filtering is performed at the same time, and the amplified electromyographic signal is full-wave rectified, and the processed data is transmitted to the single-chip microcomputer. A micro vibration sensor is added to the end of the fingertip, and the electroencephalogram sensor is used to collect the human brain electrical signal, such as Fig.17 As shown, the concentration information is extracted, the EEG signals are amplified and filtered, and the processed data is sent to the microcontroller to control the exoskeleton. This part collects the signals of each sensor and combines them with the control algorithm to control the movement of the exoskeleton. In addition, the use of vibration for grip force feedback can not only allow patients to grasp the real-time grip status, but also help remind patients to concentrate and improve their grip willingness, indirectly causing changes in EEG and EMG signals, and enhancing grip control signals, which can effectively improve the training effect and realize two-way feedback from the exoskeleton to the person and from the person to the exoskeleton.

[0015] Furthermore, the control of the hand exoskeleton robot needs to ensure a high degree of safety. Traditional position control is difficult to control the contact force within a safe range when encountering external disturbances such as collisions and blockages, which may cause personal injury or mechanical damage. To this end, this design studies the trajectory tracking control of the exoskeleton under unknown dynamic models, external disturbances, friction, and various physical constraints. In order to ensure the accurate estimation of unknown variables and improve the tracking performance, a full-actuation (FAS) composite control strategy is proposed. The algorithm can reduce the impact of external disturbances and enable the robot to respond adaptively to external forces, thereby protecting the structure of the robot and ensuring the safety of the wearer. The control combines the inner loop of the high-order disturbance observer and the outer loop of the FAS trajectory tracking controller. Specifically, the high-order disturbance observer is used to deal with uncertain dynamics and external disturbances. In addition, the controller gain is optimized using the gradient-based optimal parameter tuning method. Impedance control is characterized by rigid and precise response characteristics, and adjusts the robot's force based on position or motion control. It is suitable for scenarios where precise force or position control needs to be applied to the external environment. The data required for impedance control include reference joint angle, current joint angle, gripping force, and impedance parameters. The reference joint angle is determined by the electromyographic signal; the current joint angle is measured by the angle potentiometer; the gripping force is indirectly obtained by calculating the spring deformation through the electric push rod encoder and the angle potentiometer. The impedance parameter is set by the EEG signal. The control process is as follows: Fig.13 shown.

[0016] The system identification formula is as follows:

[0017]

[0018] Among them, f i represents the i-th element of f, F Ci and F Vi Denote the Coulomb friction coefficient and the viscous friction coefficient, B i Represents the bias term.

[0019] The design formula of the high-order disturbance observer is as follows:

[0020]

[0021] in, Represents τ d (i―1) The estimated value of , τ≈(i-1)d is the estimation error. When i=1, τ d is τ d The estimated value of .

[0022] The control rate formula of FAS compound controller is as follows:

[0023]

[0024] Among them, the gain K is selected 0 and K 1 To ensure closed-loop stability, the tracking error can be made arbitrarily small by appropriately adjusting the control gain.

[0025] The parametric design formula is as follows:

[0026] A(K 0～1 )=VFV ―1 ,detV≠0

[0027] The formula of the gradient optimization algorithm is as follows:

[0028]

[0029] The equation for impedance control is as follows:

[0030]

[0031] Where F e is the gripping force, obtained indirectly from the spring deformation; θ r is the reference angle, obtained by electromyographic signal processing; θ is the target angle, obtained by calculation. Among them, θ is guaranteed by the position loop. When F e =0,θ=θ r , the target angle coincides with the reference angle, and the effect is the same as position control. e When K≠0, the target angle deviates from the reference position as the external force moves, thus achieving the purpose of precise control. d ,B d is the impedance parameter, which affects the stiffness and damping of the system respectively. Adjusting the impedance parameter can change the stiffness of the system. The initial impedance parameter is determined by experiment. The exoskeleton robot is designed to have a small movement speed and acceleration, so K d The concentration signal obtained by the EEG sensor can have the greatest impact on K d Fine-tune so that when the concentration is high, K d When the concentration is low, K d decreases, and the system stiffness decreases.

[0032] Furthermore, this design uses electromyographic sensors to measure and process electromyographic signals to obtain the activation level of the muscles, and establishes a mapping relationship between the muscle activation level and the joint angle, and uses an impedance controller to control the joint angle; at the same time, the concentration signal collected by the EEG sensor is used to fine-tune the impedance parameters, thereby adjusting the stiffness and gripping force of the robot to achieve the effect of rehabilitation training. In order to extract effective signals and filter out high and low frequency noise, a 10-500Hz bandpass filter is used; in order to eliminate power frequency interference, a 50Hz notch filter is used; in order to make the mean value of the signal zero for subsequent processing, it is necessary to filter out its DC value; through full-wave rectification and a 1Hz low-pass filter, the muscle activation level (muscle strength) can be obtained. Two control schemes are given later, where E1 and E2 represent the muscle strength of flexors and extensors respectively; k1, k2, k3 and T are constants determined by experiments.

[0033] 1. Use muscle strength to control the speed of grasping movement. The movement speed of the fingers is determined by the strength of the muscles. The greater the muscle strength, the greater the movement speed. The advantage is that it is highly controllable and can easily control the angle at which the joint stays.

[0034] 2. Use muscle strength to control the degree of opening and closing of the palm. The angle of the joint is determined by the strength of the muscle. The greater the muscle strength, the greater the range of motion. Its advantage is quick response and sensitive control.

[0035] Technical effects and advantages of the present invention:

[0036] (1) According to the design of coupled motion of multiple four-bar linkages, the position of the adjustment slider on the wearable component can be adjusted to adapt to different hand sizes, achieve individual differentiation and adaptability, meet the needs of different human bodies, and at the same time, meet the one-to-one correspondence between the exoskeleton motion center and the human joint center. After adjusting the wearable component, the kinematic inverse solution can be used to obtain the connecting rod length according to the change of the initial position potentiometer angle, and perform kinematic analysis.

[0037] (2) According to the movement law of human fingers, the structure of the four-bar coupling mechanism is optimized to make the PIP and DIP joints closer. The angle relationship satisfies the coordination of human hands during exercise and increases comfort.

[0038] (3) Based on the displacement of the rotary potentiometer and the linear push rod, the compression of the SEA series elastic actuator spring can be calculated, the tensile force can be estimated, and force feedback without force sensors can be realized, thus reducing the number of sensors, the size of the exoskeleton system, and the overall mass.

[0039] (4) The control is achieved by integrating EEG signals and electromyographic signals, which can achieve a high level of human-computer interaction and compatibility between the human body and the exoskeleton, and realize human-to-exoskeleton feedback; at the same time, a vibration sensor is used at the end of the fingertip to transmit information to the human body through vibration, which is used to feedback the patient's grip force. When the grip force is insufficient, the vibration frequency is high and the amplitude is large, so as to promote the patient's concentration and increase the willingness to grasp, realize the exoskeleton's feedback to the person, and realize the two-way feedback of the exoskeleton to the person and the person to the exoskeleton, which has a higher level of human-computer interaction.

[0040] (5) A full-drive (FAS) composite control strategy is proposed to eliminate unnecessary system dynamics and directly assign any required closed-loop system dynamics. The proposed control strategy can not only identify unknown dynamic models, but also effectively suppress friction and external interference. A gradient-based parameter optimization method is proposed. The optimized control gains meet the joint angle limit, joint speed limit, and input torque limit. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 Schematic diagram of the hand exoskeleton

[0042] Figure 2 Schematic diagram of the index finger of the hand exoskeleton

[0043] Figure 3 Schematic diagram of the explosion of the index finger of the hand exoskeleton

[0044] Figure 4 Schematic diagram of the hand exoskeleton thumb

[0045] Figure 5 Schematic diagram of hand exoskeleton thumb explosion

[0046] Figure 6 SEA driver module schematic

[0047] Figure 7 Diagram of adjusting the wearable module

[0048] Figure 8 Vibration module schematic

[0049] Fig. 9 Schematic diagram of the tension and pressure estimation method

[0050] Fig.10 Schematic diagram of stiffness scale

[0051] Fig.11 SEA calibration result diagram

[0052] Fig.12 FAS control logic flow chart

[0053] Fig.13 Control strategy overview diagram

[0054] Fig.14 EMG signal control flow chart

[0055] Fig.15 EMG signal processing

[0056] Fig.16 Myoelectric wristband

[0057] Fig.17 EEG headband

[0058] In the figure: 1-fixed connecting rod 2-fixed straight rod coupling connecting rod 3-MCP coupling connecting rod 4-slider connecting piece 5-proximal knuckle wearing component 6-PIP driving connecting rod 7-PIP adjusting slider 8-D IP adjusting slider 9-distal knuckle wearing component 10-D IP drive link 11-MCP drive link 12-base 13-linear push rod 14-push rod SEA connection module 15-SEA sleeve 16-hexagonal nut 17-linear push rod fixing frame 18-stepped shaft 19-parent nail 20-stepped shaft 21-mid-end knuckle wearing part 22-spring guide column 23-rotary potentiometer 24-stepped shaft 25-slider fixing plate 26-flange bearing 27-stepped shaft 28-circlip 29-stepped shaft 30-stepped shaft 31-SEA sleeve end cover front 32-linear bearing 33-SEA sleeve end cover rear 34-compression spring 35-SEA guide shaft 36-stepped shaft 37-bolt 38-stepped shaft 39-stepped shaft 40-thumb MCP drive link 41-stepped shaft 42-fixed link coupling link 43-thumb fixed link 44-thumb D IP driving rod 45-distal knuckle wearing part (upper plate) 46-D IP adjusting slider 47-distal knuckle wearing part (lower plate) 48-stepped shaft 49-stepped shaft 50-slider 51-stepped shaft 52-MCP coupling rod 53-stepped shaft 54-vibration sensor DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] The user wears a hand exoskeleton and connects the hand and the exoskeleton through Velcro and gloves, such as Figure 1 As shown, for the index finger, Figure 2 , Figure 3As shown, first, the linear push rod 13 is fixed to the base 12 by screws and nuts, and the end of the linear push rod 13 is connected to the fixing frame 17 by a hexagonal nut and a bolt. One end is fixed to the base 12, and the other end is connected to the SEA push rod connector 14 by a screw. The SEA push rod connector 14 is connected to the SEA guide shaft 35 through a threaded hole. The SEA guide shaft 35 is sleeved with a compression spring 34. There are two linear bearings 32 at both ends of the SEA sleeve 15. The two ends of the sleeve 15 are fixed by the front end cover 31 of the sleeve and the rear end cover 33 of the sleeve by screws. The linear bearing 32 fixes the two ends of the compression spring 34. The push rod pushes the SEA guide shaft 35 forward, compressing the compression spring 34 in the SEA sleeve 15 to move, thereby realizing smooth movement. The drive module is as shown in FIG. Figure 6 As shown. The SEA sleeve 15 is connected to the MCP drive link 11 through a stepped shaft 36, the shaft and the link are connected through a flange bearing 26, and both ends are axially fixed by a retaining spring 28 to drive the MCP drive link 11. The MCP drive link 11 is equipped with a rotary potentiometer 23. The fixed straight rod 1 and the fixed straight rod coupling link 2 are connected through a stepped shaft, the shaft and the link are connected through a flange bearing 26, and both ends are fixed by a retaining spring 28. The fixed straight rod 1 is connected to the base 12 through screws, the fixed straight rod coupling link 2 is connected to the MCP coupling link 3 through a shaft 30, the shaft and the link are connected through a flange bearing 26, and both ends are fixed by a retaining spring 28. The MCP drive link is connected to the slider connector 4 through a flange bearing 26, and the slider connector is connected to the proximal knuckle wear component 5 through a slide rail to drive the MCP joint movement and adjust the wear module as shown in FIG. Figure 7 As shown. The MCP coupling link 3 is connected to the PIP driving link through the shaft 27, the shaft and the link are connected through the flange bearing 26, and the two ends are fixed by the retaining spring 28. The PIP driving link 6 is equipped with a rotary potentiometer 23, and the PIP driving link 6 is connected to the PIP adjusting slider 7 through the shaft shoulder. The two ends are fixed by the retaining spring 28, the PIP adjusting slider 7 is connected to the distal finger joint wearing part 9, and there is a locking nut inside the adjusting slider, which is connected to the spring guide column 22, and the spring guide column 22 is connected to the PIP adjusting slider 7 through the slider fixing plate 25. The position of the slider 7 is adjusted by rotating the spring guide column 22 to drive the PIP joint to move. The P IP driving link 6 is connected to the D IP driving link 10 through the parent-child nail 19, and a rotary potentiometer 23 is installed on the driving link 10. The driving link 10 is connected to the D IP adjustment slider 8 through a stepped shaft 26. The shaft is connected to the driving link 10 through a flange bearing 26, and both ends are fixed with a retaining ring 28. The D IP adjustment slider 8 is connected to the distal finger joint wearing component 9 through a slide groove. There is a lock nut in the D IP slider 8, and the lock nut is connected to the spring guide column 22. The spring guide column 22 is fixed to the distal finger joint wearing component 9 through a slider fixing plate 25. The position of the rotating spring guide column can adjust the position of the slider 8 on the wearing component 9, and the D IP driving link 10 drives the D IP joint to move.

[0061] For the thumb, Figure 4 , Figure 5 As shown, the linear push rod 13 is connected to one end of the base 12, and the other end is connected to the SEA push rod connection module 14 through a threaded nail. The SEA push rod connection module 14 is connected to the SEA guide shaft 35 through a threaded hole. The guide shaft 35 is connected to the SEA sleeve 15. There is a linear bearing 32 at each end of the sleeve 15. There is a compression spring 34 at the shoulder between the linear bearing 32 and the SEA guide shaft 35. There is a certain pressure in the initial state. The two ends of the sleeve are fixed by screws through the front end cover 31 of the sleeve and the rear end cover 33 of the sleeve. The compression spring is sleeved on the SEA guide shaft 35. The linear push rod 13 pushes the SEA guide shaft 35, so that the compression spring 34 is compressed, driving the stepped shaft 39 to move. The shaft is connected to the MCP drive connecting rod 40 through a flange bearing 26, and the two ends are connected by a retaining spring 28. The MCP driving connecting rod 40 is connected to the fixed straight rod 43 through the stepped shaft 41, the shaft and the connecting rod are connected through the flange bearing 26, and the two ends are fixed with the retaining spring 28. The fixed straight rod 43 is connected to the base 12 through screws and fixed on the base. The MCP coupling connecting rod 52 is connected to the fixed straight rod 43 through the shaft 53, the shaft and the connecting rod are connected through the flange bearing 26, and the two ends are fixed with the retaining spring 28. The MCP coupling connecting rod is connected to the MCP driving connecting rod 40 through the shaft 51, the shaft and the driving connecting rod 40 are connected through the flange bearing 26, and the two ends of the shaft are fixed with the retaining spring 28. The MCP driving connecting rod 40 is connected to the slider connector 4 through the flange bearing 26, and the slider connector is connected to the proximal finger joint wearing component through the slide rail 50 to drive the MCP joint movement. The MCP coupling connecting rod 52 and D The IP driving link 44 is connected by the shaft 49, the shaft is connected to the D IP driving link by the flange bearing 26, the two ends of the shaft are fixed by the retaining spring 28, the driving link 44 is equipped with a rotary potentiometer 23, the D IP driving link is connected to the adjustment slider 46 by the shaft 48, the shaft is connected to the driving link by the flange bearing 26, the two ends are fixed by the retaining spring 28, the adjustment slider 46 is connected to the distal knuckle wearing part (upper plate) 45 by the slide groove, the slider 46 has a locking nut, and is connected to the spring guide column 22, the spring guide column 22 is connected to the wearing part (upper plate) 45 by the slider fixing plate, and the position of the slider 46 can be adjusted by rotating the spring guide column 22. The distal knuckle wearing part (upper plate) 45 is connected to the distal knuckle wearing part (lower plate) 47, and the Velcro is passed through the middle, and the D IP driving link 40 drives the D IP joint to move.

[0062] The angle information of the finger is measured by the rotary potentiometer 13, and the fingertip force signal is measured by the vibration sensor 12. Figure 8As shown, the pulling pressure of the push rod is obtained by the rotary potentiometer 13 and the displacement of the push rod, and the EEG signal is obtained by the EEG sensor in the EEG headband. The original EEG data is measured to propose the concentration. First, the healthy side or the rehabilitation therapist wears an EMG bracelet to collect EMG signals, perform posture recognition and muscle strength estimation on the EMG signals, control the exoskeleton through a single-chip microcomputer, and add an impedance control algorithm to control the joint angle using an impedance controller; at the same time, the concentration signal collected by the EEG sensor is used to fine-tune the impedance parameter, thereby adjusting the stiffness and grip of the robot. By wearing an EEG headband to collect EEG signals, extract concentration, and watch the GUI interface, when the concentration is high, the impedance parameter is small and the stiffness is small; when the concentration is low, the impedance parameter is large and the stiffness is large. At the same time, the vibration signal can be fed back to the person. When the grip is insufficient or the concentration is low, the fingertips will vibrate to remind the subject to concentrate on rehabilitation. The uncertain dynamics and external disturbances are handled by combining the inner loop of the high-order disturbance observer and the outer loop of the FAS trajectory tracking controller. The gradient-based optimal parameter tuning method ensures accurate estimation of unknown variables and improves tracking performance.

[0063] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0064] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. An underactuated hand rehabilitation exoskeleton robot, comprising a connecting rod module (1), wherein the connecting rod module is equipped with an adjustable wearable module (4), characterized in that: A SEA series elastic driver (32) is installed at the end of the connecting rod module (1), a vibration sensor (54) is installed in the tip finger sleeve (9) of the connecting rod module (1), and an angle sensor (23) is embedded on the connecting rod module (1).

2. The underactuated hand rehabilitation exoskeleton robot according to claim 1, characterized in that: The connecting rod module (1) uses the human finger joint as a rigid connecting rod to participate in the connecting rod movement. The connecting rod mechanism (1) is composed of a plurality of connecting rods, the connecting rods are connected by shafts (36), and the shafts and connecting rods are connected by flange bearings (26) and pins (19).

3. The underactuated hand rehabilitation exoskeleton robot according to claim 1, characterized in that: The adjustable wearable module (4) comprises a spring guide column (22), which is clamped in a fixing plate (25), a self-locking nut is installed on the spring guide column (22), and the self-locking nut is clamped in an adjustment module (8), the adjustment module (8) is connected to a shaft (24) via a flange bearing (26), and the adjustment module (8) is mounted on a wearable component (9).

4. The underactuated hand rehabilitation exoskeleton robot according to claim 1, characterized in that: The SEA series elastic driver (32) is composed of two compression springs (34) connected in series. The compression spring (34) is wound on the SEA guide shaft (58). The guide shaft (58) is connected by two linear bearings (32). The linear bearings (32) are installed in the sleeve (15). The end of the SEA guide shaft (58) is connected to the linear push rod (13).

5. The underactuated hand rehabilitation exoskeleton robot according to claim 1, characterized in that: The vibration sensor (54) collects vibration signals, and the angle sensor (23) collects joint angle signals.

6. The underactuated hand rehabilitation exoskeleton robot according to claim 1, characterized in that: The angle signal collected by the angle sensor (23) is combined with the data of the encoder in the linear actuator (13) to realize force feedback without force sensor.

7. An underactuated hand rehabilitation exoskeleton robot, comprising an electroencephalogram sensor and an electromyography sensor.

8. The underactuated hand rehabilitation exoskeleton robot according to claim 7, characterized in that: The EEG sensor collects EEG signals and extracts concentration information, and the EMG sensor collects EMG signals and extracts posture information and muscle strength information.

9. A full-drive (FAS) composite impedance control algorithm for brain and myoelectric fusion, characterized by: Use electromyographic signals for posture recognition and muscle force estimation, and use impedance controllers to control joint angles; The concentration signals collected by the EEG sensor are used to fine-tune the impedance parameters and adjust the stiffness and gripping force of the robot. The fingertip vibration signals are used to remind the subjects to improve their concentration in training and increase their willingness to actively grasp the rehabilitation training. The high-order disturbance observer inner loop and the FAS trajectory tracking controller outer loop are combined to handle uncertain dynamics and external disturbances, and the gradient-based optimal parameter setting method is used to ensure accurate estimation of unknown variables and improve tracking performance.

Images