A control method for automatic gait adjustment in the event of a stumbling exoskeleton

Abstract

This invention belongs to the field of robotic control, specifically a control method for automatically adjusting the gait of an exoskeleton in the event of a stumble. The method comprises the following steps: acquiring wearer data collected by encoders and pressure sensors; determining the wearer's motion state based on the wearer's data; selecting a control mode based on the wearer's motion state; and calculating different driver control quantities based on different control modes to control the drive motor to control the exoskeleton robot. The present invention utilizes joint encoders, plantar pressure sensors, and toe pressure sensors to collect motion information. The designed controller is capable of driving the exoskeleton to lift its foot over an obstacle if the wearer stumbles, achieving self-balancing in the event of a stumble.

Description

Technical Field

[0001] The invention belongs to the field of robot control, in particular to a control method for automatic gait adjustment of an exoskeleton in the event of a stumbling situation. Background Art

[0002] An exoskeleton is a wearable device that can assist human movement and improve the wearer's athletic ability. Currently, there are a variety of exoskeleton control methods that drive the exoskeleton to assist the wearer. However, most of the existing control methods are only applicable to terrain without obstacles, and when tripped, they cannot regain balance on their own and fall. For example, an exoskeleton control method described in the invention patent application with application number CN201910981434.7 cannot solve the situation when the wearer is tripped by an obstacle. The invention patent application with application number CN202011497152.9 collects information through infrared devices, photoelectric sensors and IMUs to assist the wearer in movement, but it does not describe the control method when being tripped. Summary of the Invention

[0003] The purpose of the present invention is to provide an exoskeleton control method. When a person wearing an exoskeleton exercises and is tripped by an obstacle, the exoskeleton robot can automatically restore its balance to prevent the person and the robot from falling.

[0004] The technical solution adopted by the present invention to achieve the above-mentioned purpose is:

[0005] A control device for automatically adjusting the gait of an exoskeleton in the event of a stumble. Encoders are installed at the left and right hip joints and left and right knee joints of the exoskeleton robot to collect the angle values ​​and angular velocity values ​​of the left and right hip joints and the angle values ​​and angular velocity values ​​of the left and right knee joints of the wearer of the exoskeleton robot. Pressure sensors are installed at the soles and toes of the exoskeleton robot to collect the pressure values ​​of the soles and toes of the exoskeleton robot. The encoders and pressure sensors are both connected to a controller.

[0006] A control method for automatic gait adjustment in an exoskeleton stumbling situation, comprising the following steps:

[0007] Obtain wearer data collected by encoders and pressure sensors;

[0008] Determine the wearer's motion status based on the wearer's data;

[0009] Select the control mode according to the wearer's motion state;

[0010] Different driver control quantities are calculated according to different control modes, and the drive motor is controlled to control the exoskeleton robot.

[0011] The wearer data includes: the left and right hip joint angle values ​​and angular velocity values, the left and right knee joint angle values ​​and angular velocity values ​​of the exoskeleton robot wearer collected by the encoder, and the plantar pressure value and toe pressure value of the exoskeleton robot collected by the pressure sensor.

[0012] The wearer's motion states include: a stationary standing state, a walking state, and a stumble recovery state.

[0013] The determination of the wearer's motion state based on the wearer data is specifically as follows:

[0014] Set the exoskeleton robot to a stationary standing state at the start;

[0015] When the angular velocity value of any joint is greater than or equal to a first threshold, the exoskeleton robot is switched to a walking state;

[0016] When the exoskeleton robot is in a walking state, if the angular velocity value of any joint is greater than or equal to a first threshold value and the toe pressure value is less than a second threshold value, the exoskeleton robot maintains the walking state; if the angular velocity values ​​of all four joints are less than the first threshold value, the exoskeleton robot switches to a stationary standing state; if the angular velocity value of any joint is greater than or equal to the first threshold value and the toe pressure value is greater than or equal to the second threshold value, the exoskeleton robot switches to a stumble recovery state;

[0017] When the exoskeleton robot is in the stumble recovery state, when the plantar pressure value is less than a third threshold, the exoskeleton robot remains in the stumble recovery state; when the plantar pressure value is greater than or equal to the third threshold, the exoskeleton robot is switched to the stationary standing state.

[0018] The control modes include: a follow-up mode and a power-assistance mode. When the wearer's motion state is a stationary standing state, the follow-up mode is selected; when the wearer's motion state is a walking state or a stumble recovery state, the power-assistance mode is selected.

[0019] In the following mode, the wearer actively moves and drives the exoskeleton robot to move synchronously. The joint motors of the exoskeleton robot output the torque required for the exoskeleton robot's own movement. In the power-assist mode, the wearer and the exoskeleton robot drive together. The wearer and the exoskeleton robot form a human-machine whole. When the driving angle of the human-machine whole joint is inconsistent with the expected angle of the human-machine whole joint, the required torque is provided by the joint motors of the exoskeleton robot.

[0020] The following mode adopts the torque PID control method, specifically:

[0021] Perform dynamic calculations on the joint angles and angular velocities collected by the encoder to obtain the actual torques of the exoskeleton robot joints, and the controller then obtains the driving torques of the exoskeleton robot joints.

[0022] The driver control quantity is obtained by calculating the joint torque difference between the actual torque of the exoskeleton robot joint and the driving torque of the exoskeleton robot joint as well as the PID control coefficient, and then the drive motor is controlled to control the exoskeleton robot.

[0023] The power assist mode adopts the impedance PID control method, specifically:

[0024] Obtain the overall joint drive angle of the human-machine interface according to the encoder;

[0025] According to the driving angle of the human-machine overall joint and the set expected angle of the human-machine overall joint, the human-machine overall joint angle difference is obtained;

[0026] Impedance control is performed on the angle difference of the human-machine joint to obtain the desired torque of the human-machine joint;

[0027] Perform dynamic calculations on the joint angle and angular velocity values ​​collected by the encoder to obtain the driving torque of the human-machine overall joint. Calculate the human-machine overall joint torque difference based on the expected torque of the human-machine overall joint and the driving torque of the human-machine overall joint.

[0028] The driver control quantity is calculated based on the overall joint torque difference between the human and the machine and the PID control coefficient, and then the drive motor is controlled to control the exoskeleton robot.

[0029] When the exoskeleton robot is in the walking state, if the toe collides with an obstacle and the value of the toe pressure sensor is greater than or equal to the second threshold, the controller detects that the exoskeleton has been tripped by the obstacle, and the controller switches the exoskeleton robot to the trip recovery state. The exoskeleton robot starts to lift its foot and restore balance. That is, within the set time, the hip joint on the swing side continues to flex and the knee joint on the swing side continues to flex, so that the foot height is raised to above the obstacle, while the angles of the hip joint and knee joint on the supporting side remain unchanged;

[0030] When the foot is lifted to a height higher than the obstacle, the controller controls the foot to move forward and gradually fall down. When the swinging side foot contacts the ground, the plantar pressure value gradually increases. When the plantar pressure value is greater than or equal to the third threshold, the exoskeleton robot is switched to a stationary standing state and adopts a follow-up control mode. At this time, the wearer stops moving and the exoskeleton robot remains in a stationary standing state, or the wearer continues to move forward and the exoskeleton robot switches to a walking state.

[0031] The present invention has the following beneficial effects and advantages:

[0032] 1. The present invention obtains the wearer's motion data through an encoder, a plantar pressure sensor, and a toe pressure sensor. Based on the wearer's motion data, it determines whether the wearer is in a stationary standing, walking, or stumbling recovery state, and then selects a corresponding control mode based on the motion state. The control modes include: follow-up mode and power-assisted mode.

[0033] 2. The present invention can detect when the exoskeleton is tripped by an obstacle through the toe pressure sensor, thereby adding a motion disorder detection function.

[0034] 3. The present invention proposes a control method that enables the exoskeleton to start lifting its feet and cross the obstacle when it is tripped by an obstacle, ultimately achieving a self-balancing function in the event of a trip. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a schematic diagram of the structure of the exoskeleton robot used in the present invention;

[0036] Figure 2 This is a flow chart of the implementation of the exoskeleton control method of the present invention;

[0037] Figure 3 Schematic diagram of the wearer's motion state switching in the exoskeleton control method of the present invention;

[0038] Figure 4 Schematic diagram of the human-machine posture in the stumble recovery state of the exoskeleton control method of the present invention;

[0039] Figure 5 Schematic diagram of the system structure of the exoskeleton control method;

[0040] Figure 6 Schematic diagram of the control flow of the exoskeleton control method of the present invention. DETAILED DESCRIPTION

[0041] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0042] The present invention provides an exoskeleton control method, such as Figure 2 Shown, including:

[0043] Obtaining wearer data collected by an encoder, a plantar pressure sensor, and a toe pressure sensor, the wearer data including: a left hip joint angle value and angular velocity value collected by the encoder, a left knee joint angle value and angular velocity value collected by the encoder, a right hip joint angle value and angular velocity value collected by the encoder, a right knee joint angle value and angular velocity value collected by the encoder, a plantar pressure value collected by the plantar pressure sensor, and a pressure value at the point where the exoskeleton toe collides with an obstacle collected by the toe pressure sensor;

[0044] Determining the wearer's motion state based on the wearer data, the wearer's motion state including: a stationary standing state, a walking state, and a stumble recovery state;

[0045] Select the corresponding control mode according to different motion states. The control modes include: follow-up mode and power-assist mode. Specifically, when in a stationary standing state, the follow-up mode is used, and when in a walking state or a stumble recovery state, the power-assist mode is used.

[0046] According to different control modes, control signals are sent to the driver to drive the motor to control the exoskeleton;

[0047] Furthermore, the wearer's motion state is determined, such as Figure 3 As shown, specifically:

[0048] Set the starting state to a stationary standing state;

[0049] When the left hip joint angular velocity is greater than or equal to a first threshold, or the left knee joint angular velocity is greater than or equal to the first threshold, or the right hip joint angular velocity is greater than or equal to the first threshold, or the right knee joint angular velocity is greater than or equal to the first threshold, switching the exoskeleton to a walking state;

[0050] When the exoskeleton is in a walking state, if the left hip joint angular velocity is greater than or equal to a first threshold, or the left knee joint angular velocity is greater than or equal to the first threshold, or the right hip joint angular velocity is greater than or equal to the first threshold, or the right knee joint angular velocity is greater than or equal to the first threshold, the exoskeleton maintains the walking state; otherwise, it switches to a stationary standing state.

[0051] When the exoskeleton is in a walking state, when the toe pressure value is less than a second threshold, the exoskeleton maintains the walking state; when the toe pressure value is greater than or equal to the second threshold, the exoskeleton switches to a stumble recovery state.

[0052] When the exoskeleton is in the stumble recovery state, when the plantar pressure is less than a third threshold, the stumble recovery state is maintained; when the plantar pressure value is greater than or equal to the third threshold, the exoskeleton is switched to the stationary standing state.

[0053] 1) Follow-up mode

[0054] In the following motion mode, the human body actively moves and drives the exoskeleton to move synchronously, and the exoskeleton joint motors output the torque required for the exoskeleton's own movement.

[0055] Furthermore, when the human body drives the exoskeleton to move, the actual torque of the exoskeleton joints is calculated by dynamics. The controller controls the exoskeleton joint driving torque to make the exoskeleton joint driving torque close to the actual torque of the exoskeleton joints.

[0056] Specifically, the following mode adopts the torque PID control method:

[0057] The exoskeleton joint torque difference is calculated based on the actual torque of the exoskeleton joint and the exoskeleton joint driving torque, wherein the actual torque of the exoskeleton joint is obtained by performing dynamic calculation after encoder measurement data.

[0058] The driver control quantity is calculated based on the exoskeleton joint torque difference and the PID control coefficient, wherein the PID control coefficient is manually given and the driver control quantity is the exoskeleton joint driving torque increment.

[0059] The motor is driven according to the driver control quantity to control the exoskeleton.

[0060] 2) Power-Assist Mode

[0061] In the power-assisted mode, the human body and the exoskeleton are driven together, forming a human-machine whole. When the driving angle of the human-machine whole joint is inconsistent with the expected angle of the human-machine whole joint, the exoskeleton joint motor provides the required torque.

[0062] Furthermore, the desired angle of the human-machine joint is manually set and stored in the controller. When the human-machine moves, the controller calls the desired angle and compares it with the human-machine joint drive angle to obtain the human-machine joint angle difference. Using the impedance PID control method, the controller calculates the human-machine joint drive torque increment to drive the motor to move.

[0063] Specifically, the power assist mode adopts the impedance PID control method:

[0064] The human-machine overall joint angle difference is calculated according to the human-machine overall joint desired angle and the human-machine overall joint driving angle, wherein the human-machine overall joint driving angle is measured by an encoder.

[0065] Impedance control is performed according to the human-machine overall joint angle difference, and the expected torque of the human-machine overall joint is calculated, wherein the impedance control stiffness coefficient and the impedance control damping coefficient are given artificially.

[0066] The difference in torque between the human-machine joint and the human-machine joint is calculated based on the desired torque of the human-machine joint and the driving torque of the human-machine joint. The driving torque of the human-machine joint is obtained by performing dynamic calculation after encoder measurement data.

[0067] The driver control quantity is calculated based on the joint torque difference between the human and the machine and the PID control coefficient, where the PID control coefficient is manually given. The driver control quantity is the incremental torque of the joint drive of the human and the machine.

[0068] The driver control quantity is used to drive the motor to control the exoskeleton.

[0069] Furthermore, in the static standing state, the exoskeleton is in a follow-up mode.

[0070] Furthermore, in the flat ground walking state, the exoskeleton is in a power-assisted mode, and the desired angle of the exoskeleton is manually given and stored in the controller, and is called by the controller when the exoskeleton moves.

[0071] Furthermore, in the stumble recovery state, the exoskeleton is in the power-assistance mode, the exoskeleton desired angle is given manually and stored in the controller, and is called by the controller when the exoskeleton moves. The running posture is as follows Figure 4 When the exoskeleton is in the walking state, if the toe collides with an obstacle and the toe pressure sensor value is greater than or equal to the second threshold, the controller detects that the exoskeleton has tripped the obstacle and switches the exoskeleton to the trip recovery state. The exoskeleton begins to lift the foot and restore balance: within t1 seconds, the swing-side hip joint continues to flex and the swing-side knee joint continues to flex, raising the foot to a height above the obstacle, while the supporting-side hip and knee joint angles remain unchanged.

[0072] Specifically, the time t1 second for the swing side hip joint and knee joint to continue flexing movement is given manually, and the angles for the swing side hip joint to continue flexing movement and the swing side knee joint to continue flexing movement are also given manually, stored in the controller, and called by the controller during movement.

[0073] Furthermore, when the foot is lifted to a sufficient height, the controller calls the gait plan specified by humans, controls the foot to move forward, and gradually falls down. When the swinging side foot contacts the ground, the plantar pressure value gradually increases. When the plantar pressure value is greater than or equal to the third threshold, the exoskeleton switches to a stationary standing state and adopts a follow-up control mode. At this time, the wearer can stop moving and the exoskeleton remains in a stationary standing state. The wearer can also continue to move forward and the exoskeleton switches to a walking state.

[0074] Example 1:

[0075] Embodiment 1 of the present invention provides an exoskeleton control method.

[0076] Attachment Figure 1The exoskeleton robot used in the present invention is designed with straps on the thighs and calves, which can fix the lower limbs of the human body to the legs of the exoskeleton robot. The hip joints and knee joints on both sides of the exoskeleton robot are equipped with DC motors, and the motors are equipped with encoders for detecting the state of motion. A pressure sensor is installed on the toes of the robot to detect the collision of the toes with obstacles and trigger the stumble recovery state. A pressure sensor is installed on the sole of the foot to detect the contact between the swinging side and the ground in the stumble recovery state and trigger the termination of the stumble recovery state. The motor can be driven by the control method provided by the present invention, and when the wearer is tripped, a self-balancing function is provided in the event of a trip. An exoskeleton falls within the protection scope of this embodiment as long as it can be controlled using an exoskeleton control method described in any one of the present embodiments.

[0077] Attachment Figure 2 The following is a flow chart of the implementation of the exoskeleton control method, which includes the following steps:

[0078] S1: Obtain wearer data collected by the encoder, plantar pressure sensor and toe pressure sensor,

[0079] In this embodiment, the wearer data includes: the left hip joint angle value and angular velocity value collected by the encoder, the left knee joint angle value and angular velocity value collected by the encoder, the right hip joint angle value and angular velocity value collected by the encoder, the right knee joint angle value and angular velocity value collected by the encoder, the plantar pressure value collected by the plantar pressure sensor, and the pressure value at the collision point between the exoskeleton toe and the obstacle collected by the toe pressure sensor;

[0080] S2: determining the wearer's motion state according to the wearer data, specifically, the wearer's motion state includes: a stationary standing state, a walking state, and a stumble recovery state;

[0081] S3: Selecting a corresponding control mode according to different motion states. The control modes include: follow-up mode and power-assist mode. Specifically, when in a stationary standing state, the follow-up mode is used, and when in a walking state or a stumble recovery state, the power-assist mode is used.

[0082] S4: Sending control signals to the driver motors to control the exoskeleton according to different control modes;

[0083] Figure 3 A schematic diagram of the wearer's motion state switching in this embodiment, i.e., obtaining the wearer's motion state based on the wearer's motion data. As can be seen from the figure, the initial state is set to a stationary standing state;

[0084] When the left hip joint angular velocity value is greater than or equal to the first threshold value f1, or the left knee joint angular velocity is greater than or equal to the first threshold value f1, or the right hip joint angular velocity is greater than or equal to the first threshold value f1, or the right knee joint angular velocity is greater than or equal to the first threshold value f1, the exoskeleton is switched to the walking state; when the exoskeleton is in the walking state, when the left hip joint angular velocity value is greater than or equal to the first threshold value f1, or the left knee joint angular velocity is greater than or equal to the first threshold value f1, or the right hip joint angular velocity is greater than or equal to the first threshold value f1, the exoskeleton maintains the walking state, otherwise the exoskeleton is switched to the stationary standing state; when the exoskeleton is in the walking state, when the toe pressure value is less than the second threshold value f2, the exoskeleton maintains the walking state, and when the toe pressure value is greater than or equal to the second threshold value f2, the exoskeleton is switched to the stumble recovery state; when the exoskeleton is in the stumble recovery state, when the plantar pressure is less than the third threshold value f3, the stumble recovery state is maintained, and when the plantar pressure value is greater than or equal to the third threshold value f3, the exoskeleton is switched to the stationary standing state.

[0085] In this embodiment, the first to third thresholds f1 to f3 are values ​​acquired based on actual motion data. These values ​​reflect the different values ​​captured by the encoder, toe pressure sensor, and plantar pressure sensor when the wearer is in different motion states. The wearer's motion state is determined based on these values. For example, the first threshold f1 can be a value between 1° / s and 5° / s. This means that when the wearer's hip or knee pressure is greater than or equal to the first threshold f1, the wearer's motion state is switched to walking. The second threshold f2 can be a value between 1N and 10N. This means that when the wearer's toe pressure is greater than or equal to the second threshold f2, the wearer's motion state is switched to stumble recovery. The third threshold f3 can be a value between 1% and 30% of the wearer-machine total weight. This means that when the wearer's swing-side plantar pressure is greater than or equal to the third threshold f3 during the stumble recovery state, the wearer's motion state is switched to standing still, entering follow-up control mode. The above values ​​are merely illustrative and not limiting. Any value that can be used to determine the wearer's motion state through the above determination process can be used as the threshold value of this embodiment.

[0086] In step S3, the control modes include: a follow-up mode and a power-assist mode. The specific two control modes are described as follows.

[0087] 1) Follow-up mode

[0088] In the following motion mode, the human body actively moves and drives the exoskeleton to move synchronously, and the exoskeleton joint motors output the torque required for the exoskeleton's own movement.

[0089] Furthermore, when the human body drives the exoskeleton to move, the actual torque of the exoskeleton joints is calculated by dynamics, and the exoskeleton joint driving torque is controlled so that the exoskeleton joint driving torque is close to the actual torque of the exoskeleton joints.

[0090] Specifically, the following mode adopts the torque PID control method:

[0091] The exoskeleton joint torque difference is calculated based on the actual exoskeleton joint torque and the exoskeleton joint driving torque, which is expressed as:

[0092] e ET =T Ea -T Ed (1)

[0093] Among them, e ET represents the exoskeleton joint torque difference, T Ea It represents the actual torque of the exoskeleton joint, which is obtained by dynamic calculation after encoder measurement data. Ed Represents the exoskeleton joint driving torque.

[0094] Combined exoskeleton joint torque difference e ET The drive control quantity is calculated by the PID control coefficient and expressed as:

[0095]

[0096] Among them, ∫e ET dt represents the integral of the exoskeleton joint torque difference, represents the differential of the exoskeleton joint torque difference, k p 、k i 、k d represents the PID control coefficient, which is the proportional, integral, and differential coefficient of PID control, respectively. They can all be obtained by fitting experimental data. u represents the driver control quantity, which is the exoskeleton joint driving torque increment ΔT here. Ed .

[0097] The driver control quantity is used to drive the motor to control the exoskeleton.

[0098] 2) Power-Assist Mode

[0099] In the power-assisted mode, the human body and the exoskeleton are driven together, forming a human-machine whole. When the driving angle of the human-machine whole joint is inconsistent with the expected angle of the human-machine whole joint, the exoskeleton joint motor provides the required torque.

[0100] Furthermore, the desired angle of the human-machine joint is manually set and stored in the controller. When the human-machine moves, the controller calls the desired angle and compares it with the human-machine joint drive angle to obtain the human-machine joint angle difference. Using the impedance PID control method, the controller calculates the human-machine joint drive torque increment to drive the motor to move.

[0101] Specifically, the power assist mode adopts the impedance PID control method:

[0102] The human-machine overall joint angle difference is calculated based on the expected human-machine overall joint angle and the human-machine overall joint drive angle, which is expressed as:

[0103] e Mq =q Ms -q Md (3)

[0104] Among them, e Mq Indicates the overall joint angle difference between human and machine, q Ms Indicates the desired angle of the human-machine joint, which is manually specified and stored in the controller. Md Indicates the overall joint drive angle of the human-machine interface, measured by the encoder.

[0105] According to the overall joint angle difference between human and machine, e Mq Perform impedance control and calculate the expected torque of the human-machine joint as a whole, which is expressed as:

[0106]

[0107] Among them, T Ms represents the expected torque of the human-machine joint as a whole, represents the differential of the overall joint angle difference between the human and the machine, a i represents the impedance control stiffness coefficient, b i The impedance controlled damping coefficient, impedance controlled stiffness coefficient and impedance controlled damping coefficient can be obtained by fitting the experimental data.

[0108] According to the expected torque T of the human-machine joint Ms The difference between the human-machine joint torque and the human-machine joint driving torque is calculated as follows:

[0109] e MT =T Ms -T Md (5)

[0110] Among them, e MT Indicates the overall joint torque difference between human and machine, T Md It represents the driving torque of the human-machine joint as a whole, which is obtained by dynamic calculation after encoder measurement data.

[0111] According to the overall joint torque difference between human and machine e MT And PID control coefficient, calculate the drive control quantity, expressed as:

[0112]

[0113] Among them, ∫e MT dt represents the integral of the joint torque difference between the human and the machine. represents the differential of the overall joint torque difference between the human and the machine, k p 、k i 、k d Represents the PID control coefficient, which is the proportional, integral, and differential coefficient of PID control, and can be obtained by fitting experimental data. u represents the driver control quantity, which is the increment of the driving torque of the human-machine joint ΔT. Md .

[0114] The driver control quantity is used to drive the motor to control the exoskeleton.

[0115] Attachment Figure 4 Schematic diagram of the human-machine posture in the stumble recovery state of the exoskeleton control method.

[0116] When the exoskeleton is walking, it is in power-assisted mode. The desired angles of the human-machine joints are manually set, stored in the controller, and called upon by the controller during movement. When the toe collides with an obstacle and the toe pressure sensor value is greater than or equal to the second threshold f2, the controller detects that the exoskeleton has tripped and switches the exoskeleton to the trip recovery state, where it begins to lift its foot and regain balance.

[0117] Assume that the time when the stumble recovery state begins is t0, and the flexion angle of the hip joint on the swing side is θ h0 , the swing side knee flexion angle is θ k0 During t1 seconds, the hip joint on the swinging side continues to flex, with an angle of θ h1 , the knee joint on the swing side continues to flex, and the movement angle is θ k1 , so that the foot height is raised above the obstacle, while the angles of the hip and knee joints on the supporting side remain unchanged. Let this time be t2:

[0118] t2=t0+t1

[0119] At t2, the swing side hip flexion angle is θ h2 :

[0120] θ h2 =θ h0 +θ h1

[0121] At t2, the flexion angle of the swing side knee is θ k2:

[0122] θ k2 =θ k0 +θ k1

[0123] The starting time t0 of the stumble recovery state is calibrated by the controller and can be set to 0s. The swing side hip flexion angle θ h0 and the swing side knee flexion angle θ k0 The time t1 second for the hip and knee joints on the swinging side to continue flexing is given manually, and the angle θ of the hip joint on the swinging side to continue flexing is obtained by the encoder. h1 The swing side knee joint continues to flex at an angle θ k1 Given by humans.

[0124] After the flexion movement of the hip and knee joints on the swing side is completed, at time t2, the foot has been lifted to a sufficient height, and the controller calls the manually specified gait plan to make the foot move forward and gradually fall.

[0125] At time t3, the leg swings over the obstacle.

[0126] After the swinging foot contacts the ground, the plantar pressure gradually increases. When the plantar pressure reaches or exceeds a third threshold value f3, at time t4, the exoskeleton switches to a stationary standing state, adopting follow-up control mode. If the wearer stops moving, the exoskeleton remains in the stationary standing state; if the wearer continues moving forward, the exoskeleton switches to a walking state.

[0127] Attachment Figure 5 As shown in FIG. 1 , a schematic diagram of the system structure of this embodiment includes an encoder 01, a toe pressure sensor 02, a plantar pressure sensor 03, a controller 04, a driver 05, and a motor 06. The encoder 01, the toe pressure sensor 02, and the plantar pressure sensor 03 collect wearer data in real time. The controller 04 obtains the collected data and determines the wearer's motion state, switches the wearer's motion state, and matches different control modes accordingly. When the wearer is in a stationary standing state, the exoskeleton control selects the follow-up mode. When the wearer is in a walking or stumble recovery state, the power-assist mode is selected, and the driver control quantities under different control modes are sent to the driver 05 to drive the motor 06 to move and realize closed-loop control of the exoskeleton.

[0128] like Figure 6 As shown in FIG, it is a schematic diagram of the control flow of this embodiment, firstly collecting the wearer's data and determining the wearer's motion state.

[0129] If the current state is still, follow-up control is performed, that is, according to the actual torque T of the exoskeleton joint Ea , exoskeleton joint driving torque TEd Calculate the exoskeleton joint torque difference e ET , according to the exoskeleton joint torque difference e ET The PID control coefficient is used to perform PID control, and the driver control quantity u is calculated. The motor is driven according to the driver control quantity u to realize the control of the exoskeleton.

[0130] If the current state is walking or stumbling recovery, enter the power-assistance mode, that is, according to the expected angle q of the human-machine joint Ms and the overall joint drive angle q Md Calculate the overall joint rotation angle error e of the human-machine Mq , and then perform impedance control, that is, calculate the expected torque T of the human-machine joint as a whole Ms , according to the expected torque T of the human-machine joint Ms and the overall joint driving torque T Md Calculate the overall joint torque difference between human and machine e MT , according to the overall joint torque difference between human and machine e MT The PID control coefficient is used to perform PID control, and the driver control quantity u is calculated. The motor is driven according to the driver control quantity u to realize the control of the exoskeleton.

[0131] The present invention uses joint encoders, plantar pressure sensors and toe pressure sensors to collect motion information. The designed controller can drive the exoskeleton to lift the foot and cross the obstacle when the wearer is tripped by an obstacle, thus achieving self-balancing function in the event of a trip.

Claims

1. A control method for automatically adjusting the gait of an exoskeleton in the event of a stumble, the method being implemented based on a control device for automatically adjusting the gait of an exoskeleton in the event of a stumble, the device specifically comprising: installing encoders at the left and right hip joints and the left and right knee joints of the exoskeleton robot, respectively, for collecting the angle values ​​and angular velocity values ​​of the left and right hip joints and the angle values ​​and angular velocity values ​​of the left and right knee joints of the wearer of the exoskeleton robot; installing pressure sensors at the soles and toes of the exoskeleton robot, respectively, for collecting the sole pressure values ​​and toes pressure values ​​of the exoskeleton robot; the encoders and pressure sensors are both connected to a controller, characterized in that The following steps are involved: Obtain wearer data collected by encoders and pressure sensors; Determine the wearer's motion status based on the wearer's data; Select the control mode according to the wearer's motion state; Different driver control quantities are calculated according to different control modes, and the drive motor is controlled to control the exoskeleton robot; The determination of the wearer's motion state based on the wearer data is specifically as follows: Set the exoskeleton robot to a stationary standing state at the start; When the angular velocity value of any joint is greater than or equal to a first threshold, the exoskeleton robot is switched to a walking state; When the exoskeleton robot is in a walking state, if the angular velocity value of any joint is greater than or equal to a first threshold value and the toe pressure value is less than a second threshold value, the exoskeleton robot maintains the walking state; if the angular velocity values ​​of all four joints are less than the first threshold value, the exoskeleton robot switches to a stationary standing state; if the angular velocity value of any joint is greater than or equal to the first threshold value and the toe pressure value is greater than or equal to the second threshold value, the exoskeleton robot switches to a stumble recovery state; When the exoskeleton robot is in the stumble recovery state, when the plantar pressure value is less than a third threshold, the exoskeleton robot remains in the stumble recovery state; when the plantar pressure value is greater than or equal to the third threshold, the exoskeleton robot is switched to the stationary standing state.

2. A control method for automatic gait adjustment of an exoskeleton in the event of a stumble according to claim 1, characterized in that: The wearer data includes: the left and right hip joint angle values ​​and angular velocity values, the left and right knee joint angle values ​​and angular velocity values ​​of the exoskeleton robot wearer collected by the encoder, and the plantar pressure value and toe pressure value of the exoskeleton robot collected by the pressure sensor.

3. The control method for automatic gait adjustment of an exoskeleton in the event of a stumble according to claim 1, characterized in that: The wearer's motion states include: a stationary standing state, a walking state, and a stumble recovery state.

4. A control method for automatic gait adjustment of an exoskeleton in the event of a stumble according to claim 1 or 3, characterized in that: The control modes include: a follow-up mode and a power-assistance mode. When the wearer's motion state is a stationary standing state, the follow-up mode is selected; when the wearer's motion state is a walking state or a stumble recovery state, the power-assistance mode is selected.

5. The control method for automatic gait adjustment of an exoskeleton in the event of a stumble according to claim 4, characterized in that: In the following mode, the wearer actively moves and drives the exoskeleton robot to move synchronously. The joint motors of the exoskeleton robot output the torque required for the exoskeleton robot's own movement. In the power-assist mode, the wearer and the exoskeleton robot drive together. The wearer and the exoskeleton robot form a human-machine whole. When the driving angle of the human-machine whole joint is inconsistent with the expected angle of the human-machine whole joint, the required torque is provided by the joint motors of the exoskeleton robot.

6. The control method for automatic gait adjustment of an exoskeleton in the event of a stumble according to claim 5, characterized in that: The following mode adopts the torque PID control method, specifically: Perform dynamic calculations on the joint angles and angular velocities collected by the encoder to obtain the actual torques of the exoskeleton robot joints, and the controller then obtains the driving torques of the exoskeleton robot joints. The driver control quantity is obtained by calculating the joint torque difference between the actual torque of the exoskeleton robot joint and the driving torque of the exoskeleton robot joint as well as the PID control coefficient, and then the drive motor is controlled to control the exoskeleton robot.

7. The control method for automatic gait adjustment of an exoskeleton in the event of a stumble according to claim 5, characterized in that: The power assist mode adopts the impedance PID control method, specifically: Obtain the overall joint drive angle of the human-machine interface according to the encoder; According to the driving angle of the human-machine overall joint and the set expected angle of the human-machine overall joint, the human-machine overall joint angle difference is obtained; Impedance control is performed on the angle difference of the human-machine joint to obtain the desired torque of the human-machine joint; Perform dynamic calculations on the joint angle and angular velocity values ​​collected by the encoder to obtain the driving torque of the human-machine overall joint. Calculate the human-machine overall joint torque difference based on the expected torque of the human-machine overall joint and the driving torque of the human-machine overall joint. The driver control quantity is calculated based on the overall joint torque difference between the human and the machine and the PID control coefficient, and then the drive motor is controlled to control the exoskeleton robot.

8. The control method for automatic gait adjustment of an exoskeleton in the event of a stumble according to claim 1, characterized in that: When the exoskeleton robot is in the walking state, if the toe collides with an obstacle and the value of the toe pressure sensor is greater than or equal to the second threshold, the controller detects that the exoskeleton has been tripped by the obstacle, and the controller switches the exoskeleton robot to the trip recovery state. The exoskeleton robot starts to lift its foot and restore balance. That is, within the set time, the hip joint on the swing side continues to flex and the knee joint on the swing side continues to flex, so that the foot height is raised to above the obstacle, while the angles of the hip joint and knee joint on the supporting side remain unchanged; When the foot is lifted to a height higher than the obstacle, the controller controls the foot to move forward and gradually fall down. When the swinging side foot contacts the ground, the plantar pressure value gradually increases. When the plantar pressure value is greater than or equal to the third threshold, the exoskeleton robot is switched to a stationary standing state and adopts a follow-up control mode. At this time, the wearer stops moving and the exoskeleton robot remains in a stationary standing state, or the wearer continues to move forward and the exoskeleton robot switches to a walking state.

Images