Flexible elbow joint exoskeleton rehabilitation robot with sensing and driving integrated design

Abstract

The invention discloses a sensing and driving integrated design flexible elbow joint exoskeleton rehabilitation robot which comprises a dot matrix structure flexible driver and a control system, the robot is designed based on the dot matrix structure driver, the dot matrix structure flexible driver has driving and sensing functions, and under the condition that a wearing part is not provided with a sensor, the flexible elbow joint exoskeleton rehabilitation robot can be driven by the control system. According to the rehabilitation training device, an elbow joint is assisted to do precise movement in the bending and stretching directions, limb movement information is fed back to an air pressure control box body based on a self-sensing algorithm of a dot matrix structure driver, and efficient, low-cost and stable rehabilitation training is achieved.

Description

Technical Field

[0001] The present invention belongs to the field of exoskeleton rehabilitation robots, and in particular relates to a flexible elbow joint exoskeleton rehabilitation robot with an integrated perception and drive design. Background Art

[0002] Limb disabilities caused by cardiovascular diseases like stroke severely impact patients' quality of life. Cost-effective, efficient postoperative rehabilitation training is crucial for their future well-being and work. Currently, most upper limb rehabilitation robots utilize rigid linkage structures, resulting in bulky and heavy weight. Furthermore, these robots face challenges such as manufacturing difficulties and high costs, preventing many patients from receiving effective rehabilitation training.

[0003] Flexible wearable rehabilitation robots can effectively solve the above problems. Due to the advantages of high flexibility and safety of their materials, they have attracted widespread attention.

[0004] Patent CN111184620B proposes a flexible rope-driven elbow joint exoskeleton robot with a compensation device. The power structure is placed at the rear through rope transmission, which reduces the pressure on the patient's wearing part and improves the wearability of the exoskeleton robot.

[0005] Patent CN105796286B proposes a lower limb exoskeleton robot control method using an airbag sensor. An airbag pressure sensor is added to the traditional lower limb exoskeleton human intention detection sensor. The method reflects the force between the human body and the exoskeleton by measuring the signal generated by the human thigh compressing the airbag, thereby providing feedback on the human body's movement intention and correcting deviations in the exoskeleton control algorithm.

[0006] However, there are still some problems with the above patents:

[0007] 1. Flexible exoskeleton rehabilitation robots still have many rigid wearable parts, making the entire robot bulky and facing problems such as wearability and safety;

[0008] 2. During the control process, the robot needs to install sensors to collect information about the patient's limb status, and then perform closed-loop feedback control. This not only increases the overall cost of the robot, but may also lead to a decrease in stability. Summary of the Invention

[0009] To solve the above problems, the present invention discloses a flexible elbow exoskeleton rehabilitation robot with an integrated sensing and driving design. The robot is designed based on a lattice structure driver. The lattice structure flexible driver has driving and sensing functions, and can assist the elbow joint in performing precise movements in both flexion and extension without sensors in the wearing part. The entire structure is made of pure flexible TPU material, which is small in size, light in weight, comfortable to wear and low in cost.

[0010] To achieve the above object, the technical solution of the present invention is as follows:

[0011] A flexible elbow exoskeleton rehabilitation robot with integrated sensing and driving design, including a lattice structure flexible driver and a control system;

[0012] The lattice-structured flexible actuator consists of five drive units: Drive Unit 1, Drive Unit 2, Drive Unit 3, Drive Unit 4, and Drive Unit 5. Each drive unit has a diamond-shaped structure and is connected in an L-shaped pattern. Drive Units 1 and 2 are located on the outside of the upper arm, Drive Unit 3 is located on the outside of the elbow, and Drive Units 4 and 5 are located on the outside of the upper forearm. A turning joint is designed at Drive Unit 3. The actuator is connected to the arm via Mounts 1 and 2, which are made of fabric to ensure comfortable and safe wear.

[0013] Each drive unit consists of three airbags: a flexion airbag, an extension airbag, and a sensing airbag. These are folded, with the extension airbag and sensing airbag forming a diamond-shaped outer frame. The upper half of the flexion airbag is welded to the inner wall of the extension airbag, while the lower half is welded to the inner wall of the sensing airbag.

[0014] By inflating the flexion airbag and the extension airbag separately, stepless assistance for bidirectional movement of the patient's limbs can be achieved.

[0015] The lattice-structured flexible actuator adopts an integrated design, capable of sensing motion information during movement. The sensing airbags in drive units 1, 2, 4, and 5 detect contact forces with the flexible actuator during elbow joint movement, while the sensing airbag in drive unit 3 senses the elbow's bending angle.

[0016] The lattice-structured flexible actuator is made of TPU material. Each drive unit consists of an upper layer of a stretching airbag, a lower layer of a stretching airbag, an upper layer of a flexing airbag, a lower layer of a flexing airbag, an upper layer of a sensing airbag, and a lower layer of a sensing airbag. First, a high-frequency machine is used to form a closed airbag from the two layers of TPU material of the three airbags. Then, the ends of the flexing airbag are welded to the center of the stretching airbag and the sensing airbag. Finally, the ends of the stretching airbag and the sensing airbag are welded together.

[0017] The control system of the flexible elbow exoskeleton rehabilitation robot includes a flexible actuator-based sensing method and a control box. The control box contains a control circuit board, an electrical proportional valve, and an air pump. The control circuit board includes a voltage regulator module, an air pressure control module, a communication module, and a CPU module. The control circuit board transmits information with the mobile software via the communication module and sends control instructions to the air pressure control module, which controls the air pressure inside the airbag via the electrical proportional valve. The CPU module uses AD acquisition to identify the air pressure value of the sensing airbag and inputs the air pressure value into the recognition algorithm to obtain sensory information of the patient's limb.

[0018] The perception mechanism of the lattice structure driver:

[0019] The enclosed gas inside the sensing airbag satisfies the ideal gas equation:

[0020] (p0+p a )V0=nRT (1)

[0021] Where, P0 is the initial pressure of the sensing airbag; P a is the standard atmospheric pressure; V0 is the volume of the sensing airbag cavity in its initial state. Under this assumption, the total volume of gas within the cavity remains constant, and the gas inside the three sensing airbags is sealed, with the gas flow within the cavity being static. When a force is applied to the surface of the sensing airbag, causing the cavity volume to change by ΔV, the following equation is established based on Boyle's law:

[0022]

[0023] Where, P ΔV is the changing pressure value inside the cavity, and the external load calculation formula of the sensing airbag is:

[0024] F=pS e (3)

[0025] Where, F is the external load of the sensing airbag; p is the air pressure inside the sensing airbag; S e is the effective contact area between the cavities. Substituting formula (3) into formula (2), we can obtain the calculation formula for sensing the cavity change of the airbag during the external load compression process:

[0026]

[0027] From formula (4), we can see that as the external load |F| of the sensed airbag increases, the pressure change ΔV inside the cavity also increases accordingly, resulting in the cavity pressure P ΔV This shows that the sensing airbag is a sealed air, and when the sealed gas is under pressure, the air pressure in its internal cavity changes with the external load.

[0028] Perception method of lattice structure driver:

[0029] The robotic system uses an architecture consisting of three LSTM layers to realize the patient's limb status perception, such as Figure 8 As shown in the figure, each LSTM layer contains 100 hidden units, with a time step of 50ms, corresponding to a sampling window of approximately 500ms. To reduce the risk of overfitting, a dropout layer with a random dropout rate of 0.2 is introduced after each LSTM layer. A fully connected layer is added to output the results, including the contact force of the lattice structure actuator and the overall bending angle of the actuator.

[0030] Finally, through the motion information method based on the lattice structure driver, the motion information of the patient's limbs is obtained sensorlessly, and then the elbow joint exoskeleton robot is controlled through the closed-loop control method to achieve precise bidirectional motion control.

[0031] The effective effects of the present invention are:

[0032] 1. The robot is designed based on a lattice structure driver. The lattice structure flexible driver is based on an integrated perception-drive design and has both driving and perception functions. It can assist the elbow joint in precise flexion and extension movements without sensors on the wearable part. Based on the self-perception algorithm of the lattice structure driver, the limb movement information is fed back to the pneumatic control box to achieve efficient, low-cost and stable rehabilitation training.

[0033] 2. The flexible elbow joint exoskeleton rehabilitation robot described in the present invention is made of TPU material, which is a pure flexible material as a whole. It has the advantages of small size, light weight, comfortable wearing and low cost.

[0034] 3. The flexible elbow exoskeleton rehabilitation robot described in the present invention is based on a lattice structure driver. The driver is formed by arranging drive units. The assembly of multiple identical structures has the advantages of simple processing and control. In addition, the lattice structure driver can significantly improve the output force of the flexible driver under the same volume and mass, reduce the volume of gas required, and thus increase the response speed of the flexible driver.

[0035] 4. The drive unit of the driver is based on an antagonistic structure, and the two drive cavities act on a single rotation angle at the same time. Therefore, it is possible to achieve stable control in the intermediate state between the two states and control the driver stiffness at a single angle. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of wearing the lattice structure elbow joint exoskeleton rehabilitation robot of the present invention;

[0037] Figure 2Schematic diagram of the details of the lattice structure elbow joint exoskeleton rehabilitation robot of the present invention;

[0038] Figure 3 Schematic diagram of the buckling state of the lattice structure actuator of the present invention;

[0039] Figure 4 Schematic diagram of the extended state of the lattice structure driver of the present invention;

[0040] Figure 5 is a schematic diagram of a buckling state of the drive unit according to the present invention;

[0041] Figure 6 is a schematic diagram of the driving unit of the present invention in an extended state;

[0042] Figure 7 It is a schematic diagram of the processing method of the drive unit described in the present invention.

[0043] Figure 8 It is a schematic diagram of the patient limb status perception method based on the lattice structure driver described in the present invention.

[0044] List of Figure Symbols:

[0045] 1. Upper limb arm, 2. Mounting base 1, 3. Mounting base 2, 4. Upper limb forearm, 5. Drive unit 1, 6. Drive unit 2, 7. Drive unit 3, 8. Drive unit 4, 9. Drive unit 5, 10. Drive unit stretches the airbag, 11. Drive unit flexes the upper half of the airbag, 12. Drive unit senses the airbag, 13. Drive unit senses the lower half of the airbag, 14. Drive unit stretches the upper layer of the airbag, 15. Drive unit stretches the lower layer of the airbag, 16- Drive unit flexes the upper layer of the airbag, 17. Drive unit flexes the lower layer of the airbag, 18. Drive unit senses the upper layer of the airbag, 19. Drive unit senses the lower layer of the airbag. DETAILED DESCRIPTION

[0046] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0047] As shown in the figure, the flexible elbow joint exoskeleton rehabilitation robot with an integrated perception and driving design described in the present invention includes a lattice structure flexible driver and a control system; the lattice structure driver assists the patient's arm in bidirectional movement, and based on the self-perception algorithm of the lattice structure driver, the limb movement information is fed back to the pneumatic control box, realizing the integration of perception and driving, and can accurately assist patients in stable elbow joint rehabilitation training.

[0048] The lattice-structured flexible actuator consists of drive unit 1 5, drive unit 2 6, drive unit 3 7, drive unit 4 8, and drive unit 5 9. Each drive unit is diamond-shaped and connected in sequence to form an L-shaped structure. A turning joint is designed at drive unit 3 7. The actuator is connected to the elbow joint via mounting base 1 2 and mounting base 2 3. Drive units 1 5 and 2 6 are located on the outside of the upper arm 1, drive unit 3 7 is located on the outside of the elbow, and drive units 4 8 and 5 9 are located on the outside of the upper forearm 4.

[0049] The mounting bases 2 and 3 are made of fabric material to achieve comfortable and safe wearing.

[0050] Each drive unit consists of three airbags, such as Figure 5 As shown, the bent flexing airbags 11 and 13, the stretching airbag 10 and the sensing airbag 12 are respectively, wherein the stretching airbag 10 and the sensing airbag 12 form an outer frame of a diamond structure, the upper half 11 of the flexing airbag is welded to the inner wall of the stretching airbag 10, and the lower half 13 of the flexing airbag is welded to the inner wall of the sensing airbag 12;

[0051] During use, the flexion and extension airbags are inflated separately to achieve stepless assistance for the patient's arm's bidirectional movement. The arm is flexed when the flexion airbag is inflated, and the arm is extended when the extension airbag is inflated.

[0052] The lattice-structured flexible actuator adopts an integrated design, capable of sensing motion information during movement. The sensing airbags in drive units 1 (5), 2 (6), 4 (8), and 5 (9) sense the contact force between the elbow joint and the flexible actuator, while the sensing airbag in drive unit 3 (7) senses the elbow's bending angle.

[0053] The lattice structure flexible driver is made of TPU material. The three airbags of the driving unit are all double-layer structures, that is, a single driving unit includes the upper layer of the driving unit stretching airbag, the lower layer of the driving unit stretching airbag, the upper layer of the driving unit flexing airbag, the lower layer of the driving unit flexing airbag, the upper layer of the driving unit sensing airbag and the lower layer of the driving unit sensing airbag. Figure 7 As shown, first, a high-frequency oscillator is used to weld the two layers of TPU material of the three airbags into a closed airbag. Then, the ends of the flexion airbag are welded to the middle of the expansion airbag and the sensing airbag. Finally, the ends of the expansion airbag and the sensing airbag are welded together.

[0054] The control system of the flexible elbow exoskeleton rehabilitation robot includes a flexible actuator-based sensing method and a control box. The control box contains a control circuit board, an electrical proportional valve, and an air pump. The control circuit board includes a voltage regulator module, an air pressure control module, a communication module, and a CPU module. The control circuit board transmits information with the mobile software via the communication module and sends control instructions to the air pressure control module, which controls the air pressure inside the airbag via the electrical proportional valve. The CPU module uses AD acquisition to identify the air pressure value of the sensing airbag and inputs the air pressure value into the recognition algorithm to obtain sensory information of the patient's limb.

[0055] The perception mechanism of the lattice structure driver:

[0056] The enclosed gas inside the sensing airbag satisfies the ideal gas equation:

[0057] (p0+p a )V0=nRT (1)

[0058] Where, P0 is the initial pressure of the sensing airbag; P a is the standard atmospheric pressure; V0 is the volume of the sensing airbag cavity in its initial state. Under this assumption, the total volume of gas within the cavity remains constant, and the gas inside the three sensing airbags is sealed, with the gas flow within the cavity being static. When a force is applied to the surface of the sensing airbag, causing the cavity volume to change by ΔV, the following equation is established based on Boyle's law:

[0059]

[0060] Where, P ΔV is the changing pressure value inside the cavity, and the external load calculation formula of the sensing airbag is:

[0061] F=pS e (3)

[0062] Where, F is the external load of the sensing airbag; p is the air pressure inside the sensing airbag; S e is the effective contact area between the cavities. Substituting formula (3) into formula (2), we can obtain the calculation formula for sensing the cavity change of the airbag during the external load compression process:

[0063]

[0064] From formula (4), we can see that as the external load |F| of the sensed airbag increases, the pressure change ΔV inside the cavity also increases accordingly, resulting in the cavity pressure P ΔV This shows that the sensing airbag is a sealed air, and when the sealed gas is under pressure, the air pressure in its internal cavity changes with the external load.

[0065] Perception method of lattice structure driver:

[0066] The robotic system uses an architecture consisting of three LSTM layers to realize the patient's limb status perception, such as Figure 8 As shown in the figure, each LSTM layer contains 100 hidden units, with a time step of 50ms, corresponding to a sampling window of approximately 500ms. To reduce the risk of overfitting, a dropout layer with a random dropout rate of 0.2 is introduced after each LSTM layer. A fully connected layer is added to output the results, including the contact force of the lattice structure actuator and the overall bending angle of the actuator.

[0067] Finally, through the motion information method based on the lattice structure driver, the motion information of the patient's limbs is obtained sensorlessly, and then the elbow joint exoskeleton robot is controlled through the closed-loop control method to achieve precise bidirectional motion control.

[0068] It should be noted that the above content merely illustrates the technical idea of the present invention and cannot be used to limit the scope of protection of the present invention. For ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications all fall within the scope of protection of the claims of the present invention.

Claims

1. A flexible elbow exoskeleton rehabilitation robot with integrated sensing and driving design, characterized by: Including lattice structure flexible drive and control system; The lattice structure flexible actuator consists of five drive units: drive unit 1, drive unit 2, drive unit 3, drive unit 4, and drive unit 5. Each drive unit has a diamond structure and is connected in an L-shape. Drive units 1 and 2 are set on the outside of the upper arm, drive unit 3 is set on the outside of the elbow, and drive units 4 and 5 are set on the outside of the upper forearm. A turning joint is designed at drive unit 3. The actuator is connected to the arm through mounting brackets 1 and 2. The lattice structure flexible actuator is made of TPU material. Each drive unit consists of three airbags: a flexion airbag, an extension airbag, and a sensing airbag. The flexion airbag, extension airbag, and sensing airbag are bent, and the extension airbag and sensing airbag form a diamond-shaped outer frame. The upper half of the flexion airbag is welded to the inner wall of the extension airbag, and the lower half of the flexion airbag is welded to the inner wall of the sensing airbag. By inflating the flexion airbag and the extension airbag separately, stepless assistance of the patient's limb bidirectional movement can be achieved; The lattice-structured flexible actuator adopts an integrated design, which can sense the motion information of the actuator during movement. Among them, the sensing airbags of drive units 1, 2, 4, and 5 can sense the contact force with the flexible actuator during elbow joint movement, and the sensing airbag of drive unit 3 can sense the bending angle during elbow joint movement. The control system includes a sensing method based on a flexible driver and a control box. The control box contains a control circuit board, an electric proportional valve and an air pump. The control circuit board includes a voltage stabilizing module, an air pressure control module, a communication module and a CPU module. The control circuit board transmits information with the mobile software through the communication module and sends control instructions to the air pressure control module. The air pressure control module controls the air pressure inside the airbag through the electric proportional valve. The CPU module collects and identifies the air pressure value of the sensed airbag through AD, and inputs the air pressure value into the recognition algorithm to obtain the perception information of the patient's limbs.

2. The perception-driven integrated design flexible elbow exoskeleton rehabilitation robot according to claim 1, characterized in that: A single driving unit includes a driving unit stretching airbag upper layer, a driving unit stretching airbag lower layer, a driving unit flexing airbag upper layer, a driving unit flexing airbag lower layer, a driving unit sensing airbag upper layer and a driving unit sensing airbag lower layer.

3. The perception-driven integrated design flexible elbow exoskeleton rehabilitation robot according to claim 1, characterized in that: The preparation method of a single drive unit is: First, a high-frequency frequency machine is used to make the two layers of TPU material of the three airbags into a closed airbag; then, the two ends of the flexion airbag are welded to the middle of the expansion airbag and the sensing airbag respectively; finally, the two ends of the expansion airbag and the sensing airbag are welded together.

4. The perception-driven integrated design flexible elbow exoskeleton rehabilitation robot according to claim 1, characterized in that: The first mounting base and the second mounting base are made of fabric material.

5. The perception-driven integrated design flexible elbow exoskeleton rehabilitation robot according to claim 1, characterized in that: Perception mechanism of lattice structure flexible actuator: The enclosed gas inside the sensing airbag satisfies the ideal gas equation: (p0+p a )V0=nRT (1) Where, P0 is the initial pressure of the sensing airbag; P a is the standard atmospheric pressure; V0 is the volume of the sensing airbag cavity in the initial state; under this assumption, the total amount of gas inside the cavity remains unchanged, and the internal gas of the three sensing airbags is sealed, and the gas flow inside the cavity is static; when the sensing airbag surface is subjected to force, the sensing airbag cavity changes by ΔV. Based on Boyle's law, the following equation is established for the gas inside the cavity: Where, P ΔV is the changing pressure value inside the cavity, and the external load calculation formula of the sensing airbag is: F=pS e (3) Where, F is the external load of the sensing airbag; p is the air pressure inside the sensing airbag; S e is the effective contact area between the cavities; Substituting formula (3) into formula (2), we obtain the calculation formula for sensing the cavity change during the external load compression process of the airbag: From formula (4), we can see that as the external load |F| of the sensed airbag increases, the pressure change ΔV inside the cavity also increases accordingly, resulting in the cavity pressure P ΔV Increased; it can be seen that the sensing airbag is a confined air, and when the confined gas is under pressure, the air pressure in its internal cavity changes with the external load.

6. The perception-driven integrated design flexible elbow exoskeleton rehabilitation robot according to claim 1, characterized in that: Perception method of lattice structure flexible actuator: The robotic system uses an architecture consisting of three LSTM layers to achieve patient limb status perception. Each LSTM layer contains 100 hidden units, and the time step is set to 50ms, corresponding to a sampling window of approximately 500ms. To reduce the risk of overfitting, a dropout layer with a random dropout rate of 0.2 is introduced after each LSTM layer. A fully connected layer is added to output the results, including the contact force of the lattice structure actuator and the overall bending angle of the actuator. Finally, through the motion information method based on the lattice structure driver, the motion information of the patient's limbs is obtained sensorlessly, and then the elbow joint exoskeleton robot is controlled through the closed-loop control method to achieve precise bidirectional motion control.

Images