CN112733422A - Optimal assistance prediction method for flexible exoskeleton based on Gaussian process regression - Google Patents

Optimal assistance prediction method for flexible exoskeleton based on Gaussian process regression Download PDF

Info

Publication number
CN112733422A
CN112733422A CN202011387447.0A CN202011387447A CN112733422A CN 112733422 A CN112733422 A CN 112733422A CN 202011387447 A CN202011387447 A CN 202011387447A CN 112733422 A CN112733422 A CN 112733422A
Authority
CN
China
Prior art keywords
amplitude
assistance
data set
establishing
gaussian process
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202011387447.0A
Other languages
Chinese (zh)
Inventor
孙磊
曾德添
董恩增
佟吉刚
陈鑫
李云飞
龚欣翔
李成辉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin University of Technology
Original Assignee
Tianjin University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianjin University of Technology filed Critical Tianjin University of Technology
Priority to CN202011387447.0A priority Critical patent/CN112733422A/en
Publication of CN112733422A publication Critical patent/CN112733422A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61H—PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY
    • A61H1/00—Apparatus for passive exercising; Vibrating apparatus ; Chiropractic devices, e.g. body impacting devices, external devices for briefly extending or aligning unbroken bones
    • A61H1/02—Stretching or bending or torsioning apparatus for exercising
    • A61H1/0237—Stretching or bending or torsioning apparatus for exercising for the lower limbs
    • A61H1/0255—Both knee and hip of a patient, e.g. in supine or sitting position, the feet being moved in a plane substantially parallel to the body-symmetrical-plane
    • A61H1/0262—Walking movement; Appliances for aiding disabled persons to walk
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B21/00—Exercising apparatus for developing or strengthening the muscles or joints of the body by working against a counterforce, with or without measuring devices
    • A63B21/00181—Exercising apparatus for developing or strengthening the muscles or joints of the body by working against a counterforce, with or without measuring devices comprising additional means assisting the user to overcome part of the resisting force, i.e. assisted-active exercising
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B23/00—Exercising apparatus specially adapted for particular parts of the body
    • A63B23/035—Exercising apparatus specially adapted for particular parts of the body for limbs, i.e. upper or lower limbs, e.g. simultaneously
    • A63B23/04—Exercising apparatus specially adapted for particular parts of the body for limbs, i.e. upper or lower limbs, e.g. simultaneously for lower limbs
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
    • A63B24/0003—Analysing the course of a movement or motion sequences during an exercise or trainings sequence, e.g. swing for golf or tennis
    • A63B24/0006—Computerised comparison for qualitative assessment of motion sequences or the course of a movement
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
    • A63B24/0075—Means for generating exercise programs or schemes, e.g. computerized virtual trainer, e.g. using expert databases
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
    • A63B24/0087—Electric or electronic controls for exercising apparatus of groups A63B21/00 - A63B23/00, e.g. controlling load
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/17—Mechanical parametric or variational design
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61H—PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY
    • A61H2201/00—Characteristics of apparatus not provided for in the preceding codes
    • A61H2201/16—Physical interface with patient
    • A61H2201/1602—Physical interface with patient kind of interface, e.g. head rest, knee support or lumbar support
    • A61H2201/164—Feet or leg, e.g. pedal
    • A61H2201/1642—Holding means therefor
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61H—PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY
    • A61H2201/00—Characteristics of apparatus not provided for in the preceding codes
    • A61H2201/16—Physical interface with patient
    • A61H2201/1602—Physical interface with patient kind of interface, e.g. head rest, knee support or lumbar support
    • A61H2201/165—Wearable interfaces
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61H—PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY
    • A61H2201/00—Characteristics of apparatus not provided for in the preceding codes
    • A61H2201/50—Control means thereof
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61H—PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY
    • A61H2201/00—Characteristics of apparatus not provided for in the preceding codes
    • A61H2201/50—Control means thereof
    • A61H2201/5007—Control means thereof computer controlled
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61H—PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY
    • A61H2201/00—Characteristics of apparatus not provided for in the preceding codes
    • A61H2201/50—Control means thereof
    • A61H2201/5058—Sensors or detectors
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61H—PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY
    • A61H2201/00—Characteristics of apparatus not provided for in the preceding codes
    • A61H2201/50—Control means thereof
    • A61H2201/5058—Sensors or detectors
    • A61H2201/5061—Force sensors
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61H—PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY
    • A61H2205/00—Devices for specific parts of the body
    • A61H2205/10—Leg
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
    • A63B24/0003—Analysing the course of a movement or motion sequences during an exercise or trainings sequence, e.g. swing for golf or tennis
    • A63B24/0006—Computerised comparison for qualitative assessment of motion sequences or the course of a movement
    • A63B2024/0012—Comparing movements or motion sequences with a registered reference
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
    • A63B24/0075—Means for generating exercise programs or schemes, e.g. computerized virtual trainer, e.g. using expert databases
    • A63B2024/0081—Coaching or training aspects related to a group of users
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/50—Force related parameters
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/50—Force related parameters
    • A63B2220/51—Force
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/80—Special sensors, transducers or devices therefor
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces

Abstract

A method for predicting optimal assistance of a flexible exoskeleton based on Gaussian process regression is characterized in that data sets of comfortable assistance of different people and different walking speeds are established, human body model characteristics of testers are determined, data sets (X, Y) of samples are established on the basis, after the data sets are established and trained, distribution of prediction functions is directly defined by using a Gaussian process regression mode, the predicted assistance amplitude is enabled to be in a more reasonable range, the smooth assistance to legs is really realized, and the injured bones, muscles and joints are helped to play a role in assisting rehabilitation.

Description

Optimal assistance prediction method for flexible exoskeleton based on Gaussian process regression
The technical field is as follows:
the invention belongs to the technical field of robots, relates to power-assisted parameter optimization of a gait rehabilitation flexible exoskeleton robot, and particularly relates to a flexible exoskeleton optimal power-assisted prediction method based on Gaussian process regression.
(II) background technology:
with the age, the strength of the lower limbs of the human body is gradually weakened, the skeleton of the human body begins to become brittle, the muscles begin to degenerate, the joint activity is reduced, the labor of standing and walking for a long time under a large load cannot be carried out, and great troubles are brought to the traveling and the life of people. In order to relieve leg pressure caused by aging, leg activities of people can still be performed normally, normal production, life and labor are maintained, and the gait rehabilitation flexible exoskeleton robot can be produced at will.
The gait rehabilitation flexible exoskeleton robot is a set of power assisting device worn on a human body, and monitors the force borne by a leg when the leg is lifted up and put down in real time through a tension sensor carried on the device, so that the force is fed back to a motor in time in a feedback mode, the motor controls the contraction and stretching of a flexible nylon belt, the flexible exoskeleton robot outputs different power assisting amplitude values, and the flexible power assisting of the leg is realized. However, the determination of the power assistance amplitude parameter is always plagued, and the adoption of the classical least square linear regression method can cause the problem of overfitting the power assistance amplitude parameter. Moreover, for different people, the characteristics of the human body model are not completely the same, the magnitude of the required boosting amplitude is also not completely the same, and the danger of secondary injury of partial muscles with original functions lost and abnormal stretching can be caused by overlarge and too fast pulling force; if the tension is too small or too slow, the auxiliary function of the exoskeleton cannot be embodied; the advantages of the exoskeleton can be embodied only by comfortable assistance, and the aim of assisting rehabilitation is really achieved.
Aiming at the defects of the prior art, an optimal assistance prediction method is needed at present to solve the problems of uncertainty and instability of the output assistance amplitude of the flexible exoskeleton in the using process.
(III) the invention content:
the invention aims to provide a flexible exoskeleton optimal power prediction method based on Gaussian process regression, which can overcome the defects of overlarge and too fast output power amplitude caused by a least square normal regression mode, is simple to operate and easy to realize, is beneficial to improving the safety and reliability of exoskeleton use, reasonably and effectively predicts the output power amplitude, and has more accurate result.
The technical scheme of the invention is as follows: a method for predicting optimal assistance of a flexible exoskeleton based on Gaussian process regression comprises the following steps:
(1) establishing data sets of comfortable assistance of different crowds and different pace speeds, establishing human body model characteristics of testers, and establishing data sets (X, Y) of samples on the basis;
the method for establishing the data sets of comfortable assistance of different crowds and different pace speeds in the step (1) comprises the following steps as shown in fig. 1:
firstly, leading testers with different age groups and different human body characteristics to wear exoskeleton equipment to walk on a running machine at the speed of 7.5 Km/h;
firstly, setting an initial boosting amplitude of 24N by a motor carried on the exoskeleton through driving the flexible nylon belt to contract and stretch;
in the second step, the principle of the device for contracting and stretching the flexible nylon belt is realized by controlling the rotating speed of the motor and is shown in fig. 2, the principle is that the motor and the winding drum are fixed on an aluminum bottom plate through the meshing of two steel gears, the two selected gears must meet the meshing condition, namely the modulus and the pressure angle on the reference circle of the two selected gears are respectively and correspondingly equal. The gear connected with the motor end is smaller than the gear connected with the winding drum end, and the rotation of the motor drives the gear at the motor end to rotate, so that the gear at the winding drum end is indirectly driven, and the flexible nylon belt wound on the winding drum can be contracted and stretched by the winding drum. An aluminum base plate having a motor and a drum mounted thereon is fixed to a waistband made of a webbing, and the waistband is worn in close contact with the abdomen. One end of a first section of flexible nylon belt is fixed on the winding drum through an M3 screw, the other end of the flexible nylon belt is bound on the upper fixing device, two forks at the lower end of the Y-shaped flexible nylon belt are respectively sewn on the left and right symmetrical sides of the opening of the knee nest of the knee pad, the upper end of the Y-shaped flexible nylon belt is bound on the lower fixing device, a steel connecting piece is used for connecting the two fixing devices, and a tension sensor is placed on the connecting piece and used for monitoring the force when the leg part is lifted up and put down in real time. The Y-shaped line sewing mode can increase the contact area of the flexible nylon belt and the kneepad, ensure the stable power assistance of the flexible nylon belt during contraction and stretching, and simultaneously play a role in buffering conflict and exerting force. The control of the rotating speed of the motor can be realized through PID (proportion-integration-differentiation) control, and the speed of the rotating speed of the motor drives the winding drum through the gear to realize the speed of contraction and stretching of the flexible nylon belt so as to achieve different given forces, thereby realizing different boosting amplitudes for legs;
thirdly, the comfort level of the testing personnel is scored, namely: the tester scores the comfort degree of the leg, the comfort degree score range is 0-10, and the higher the score, the more suitable the amplitude of the assistance is;
if the tester feels that the output boosting amplitude is not comfortable, namely, the evaluation time is less than 5 minutes, the rotating speed of the motor is adjusted to drive the flexible nylon belt to contract and stretch, so that the boosting amplitudes with different sizes are output, the amplitude of comfortable boosting is in a more reasonable and effective range, the leg pressure is effectively relieved, and the bones, muscles and joints are helped to recover;
if the tester is satisfied with the current output power-assisted amplitude and marks a satisfied time division of more than 8 minutes, recording the current power-assisted amplitude as A and the walking pace as V;
sixthly, changing the current walking pace V by the tester, walking on the treadmill at a higher speed or a lower speed, and repeating the step two-fifth;
seventhly, repeating the steps until a data set containing not less than 500 personal body model characteristics, walking pace obtained through experimental tests and corresponding assistance amplitude values can be established.
The step (1) of establishing the human body model characteristics of the testers specifically comprises the following steps: the information of the age, height, weight, thigh length, shank length, knee diameter, foot length, foot width, ankle height and ankle width of the test person is obtained, and the human body characteristics and the unit relationship thereof are shown in table 1:
TABLE 1 human body characteristics and unit relationship corresponding table
Characteristics of human body Unit of
Age (age) Year of old people
Height of a person mm
Body weight Kg
Thigh length mm
Length of shank mm
Diameter of knee mm
Foot length mm
Foot width mm
Height of ankle bone mm
Width of ankle bone mm
Establishing a sample data set (X, Y) on the basis of the data set; wherein X is the walking pace containing the human body model characteristics and obtained by experimental tests, and Y is the corresponding assistance amplitude;
the method for establishing the sample data set (X, Y) comprises the following steps: respectively placing the collected human body model characteristics and walking pace V obtained by experimental tests in each row of an Excel table, and placing the corresponding assistance amplitude A in another row; for example: putting the age in a first row, the height in a second row, the weight in a third row, the thigh length in a fourth row, the shank length in a fifth row, the knee diameter in a sixth row, the foot length in a seventh row, the foot width in an eighth row, the ankle height in a ninth row, the ankle width in a tenth row, the walking pace V in a eleventh row, and the corresponding power-assisted amplitude A in a twelfth row; establishing data sets according to different age groups, establishing a data set for the age group of 20-30 years, establishing a data set for the age group of 31-40 years, establishing a data set for the age group of 41-50 years, establishing a data set for the age group of 51-60 years, establishing a data set for the age group of 61-70 years, establishing a data set for the age group of 71-80 years, and establishing data sets by age groups, so that the predicted boosting amplitude is better close to the actually required boosting amplitude.
(2) After a data set is established and trained, the distribution of a prediction function is directly defined by using a Gaussian process regression mode, wherein the Gaussian process regression is composed of a mean function and a covariance function, and the mathematical expression of the Gaussian process regression is as follows:
f|X～N(m(x),K(X,X)) (1)
wherein X is { X ═ X1,x2,......xnThe walking pace obtained by human body model characteristics and experimental tests; m (x) ═ E [ f (x)]In practical applications, m (x) is often taken as 0, and y is the observed value of the sample; k (X, X) is a covariance matrix; further, an observed value y and a predicted value f can be obtained*The joint distribution of (c) is as follows:
Figure BDA0002810078000000051
in the formula, K (X, X)*)＝K(x*,X)TIs a test point x*I.e. the covariance of any one of the X data with the training set X, K (X)*,x*) Is a test point x*(ii) its own covariance;
Figure BDA0002810078000000052
is the covariance of the sequence of error values, InIs a unit vector of f*The boosting amplitude is predicted according to different human body model characteristics and walking pace;
(3) the form of the test point, i.e. the predicted value, can be obtained according to the properties of the multivariate normal distribution and the conditional distribution, as shown in formula (3):
Figure BDA0002810078000000053
(4) giving a power-assisted amplitude of the motor, wherein the power-assisted amplitude is a predicted power-assisted amplitude obtained by training a data set and optimizing the data set through Gaussian process regression analysis, which is described in the step (2), so that a tester can wear the exoskeleton equipment again to test the exoskeleton equipment on the treadmill for multiple times at different speeds, and if the tester needs to give force to the exoskeleton equipment to lift and put down the legs freely and conveniently during testing, the current power-assisted amplitude is not the optimal power-assisted amplitude, and the auxiliary rehabilitation effect on the skeleton cannot be achieved; at the moment, the assistance amplitude A in the Excel data set is continuously modified according to the feedback of the current assistance comfort level of a tester, so that the data set is more complete and has stronger universality, the improved data set is retrained and optimized through Gaussian process regression analysis, and the aim is to enable the predicted assistance amplitude to be closer to the actually required assistance amplitude; for example: the assistance amplitude A in the original data set is 30N, but at the moment, the tester feels satisfied, and if the comfortable assistance amplitude A which scores more than 8 points of satisfaction is 35N, the 30N is modified to 35N.
(5) And (3) applying the power-assisted amplitude value after the perfect optimization in the step (4) to a motor, and if a tester feels that the leg is lifted and put down, the tester does not receive the action of self given force any more but completely drives the flexible nylon belt to contract and stretch the given force, namely: the leg is lifted and put down and is not restrained by the action of artificial force any more, the predicted assistance amplitude reaches the expected value, and the assistance amplitude A in the Excel data set is not modified at the moment; only when the leg can be lifted up and put down freely, the injured leg can be helped to walk and jogging under the assistance of the exoskeleton, so that the bone, muscle and joints in the injured leg can be effectively helped to obtain a better recovery effect, and the effect of assisting rehabilitation of the exoskeleton is reflected.
The working principle of the invention is as follows: and respectively establishing comfortable assistance data sets of different crowds and different walking speeds by adopting an experimental test mode. In addition, the human body model characteristics of a tester can be established, a sample data set (X, Y) is established on the basis, uncertainty, instability and limitation caused by the absence of a data set are overcome, the safety and reliability of the exoskeleton are improved, a more accurate assistance result is obtained compared with the assistance result obtained without the data set, and the optimization of assistance parameters is realized. However, in the process of acquiring experimental data, different crowds cannot be completely covered, and the problem of overfitting of the assistance amplitude parameter can be caused by adopting a classical least square normal linear regression mode. Therefore, the method and the device adopt a Gaussian process regression mode to directly define the distribution of the prediction function, reasonably and effectively predict the magnitude of the boosting amplitude and achieve the expected use effect.
The invention has the advantages that: the distribution of the prediction function is directly defined by a Gaussian process regression mode on the premise of establishing data sets of different crowds and different pace speeds, so that the assistance amplitude parameter is in a more reasonable range, the soft assistance to the leg is really realized, and the practicability of using the flexible exoskeleton to assist the leg is reflected.
(IV) description of the drawings:
fig. 1 is a schematic flow chart of establishing corresponding assistance amplitude data sets at different step speeds in a flexible exoskeleton optimal assistance prediction method based on gaussian process regression according to the present invention.
Fig. 2 is a schematic diagram of a device for stretching and contracting a flexible nylon belt by controlling a motor in the optimal power prediction method of the flexible exoskeleton based on gaussian process regression.
Wherein, 1, 5 are aluminium system boss, 2 are aluminium system bottom plate, 3 are the motor, 4 are the aluminium system reel, 6 are M3 screw hole, 7 are flexible nylon belt, 8, 9 are fixing device, 10 are the steel connecting piece, 11 tension sensor, 12 are wearing personnel's knee.
(V) specific embodiment:
example (b): a method for predicting optimal assistance of a flexible exoskeleton based on Gaussian process regression comprises the following steps:
(1) establishing data sets of comfortable assistance of different crowds and different pace speeds, establishing human body model characteristics of testers, and establishing data sets (X, Y) of samples on the basis;
the method for establishing the data set of comfortable assistance for different people and different pace speeds comprises the following steps as shown in fig. 1:
firstly, leading testers with different age groups and different human body characteristics to wear exoskeleton equipment to walk on a running machine at the speed of 7.5 Km/h;
firstly, setting an initial boosting amplitude of 24N by a motor carried on the exoskeleton through driving the flexible nylon belt to contract and stretch;
the principle of the device for contracting and stretching the flexible nylon belt by controlling the rotating speed of the motor is shown in fig. 2, and the principle is that the motor and the winding drum are fixed on an aluminum bottom plate through the meshing of two steel gears, and the two selected gears must meet the meshing condition, namely the modulus and the pressure angle on the reference circle of the two selected gears are respectively and correspondingly equal. The gear connected with the motor end is smaller than the gear connected with the winding drum end, and the rotation of the motor drives the gear at the motor end to rotate, so that the gear at the winding drum end is indirectly driven, and the flexible nylon belt wound on the winding drum can be contracted and stretched by the winding drum. An aluminum base plate having a motor and a drum mounted thereon is fixed to a waistband made of a webbing, and the waistband is worn in close contact with the abdomen. One end of a first section of flexible nylon belt is fixed on the winding drum through an M3 screw, the other end of the flexible nylon belt is bound on the upper fixing device, two forks at the lower end of the Y-shaped flexible nylon belt are respectively sewn on the left and right symmetrical sides of the opening of the knee nest of the knee pad, the upper end of the Y-shaped flexible nylon belt is bound on the lower fixing device, a steel connecting piece is used for connecting the two fixing devices, and a tension sensor is placed on the connecting piece and used for monitoring the force when the leg part is lifted up and put down in real time. The Y-shaped line sewing mode can increase the contact area of the flexible nylon belt and the kneepad, ensure the stable power assistance of the flexible nylon belt during contraction and stretching, and simultaneously play a role in buffering conflict and exerting force. The control of the rotating speed of the motor can be realized through PID (proportion-integration-differentiation) control, and the speed of the rotating speed of the motor drives the winding drum through the gear to realize the speed of contraction and stretching of the flexible nylon belt so as to achieve different given forces, thereby realizing different boosting amplitudes for legs;
thirdly, the comfort level of the testing personnel is scored, namely: the tester scores the comfort degree of the leg, the comfort degree score range is 0-10, and the higher the score, the more suitable the amplitude of the assistance is;
if the tester feels that the output boosting amplitude is not comfortable, namely, the evaluation time is less than 5 minutes, the rotating speed of the motor is adjusted to drive the flexible nylon belt to contract and stretch, so that the boosting amplitudes with different sizes are output, the amplitude of comfortable boosting is in a more reasonable and effective range, the leg pressure is effectively relieved, and the bones, muscles and joints are helped to recover;
if the tester is satisfied with the current output power-assisted amplitude and marks a satisfied time division of more than 8 minutes, recording the current power-assisted amplitude as A and the walking pace as V;
sixthly, changing the current walking pace V by the tester, walking on the treadmill at a faster speed of 8.8Km/h (usually 8-10 Km/h) or a slower speed of 5.5Km/h (usually 4-6 Km/h), and repeating the step two to the fifth step;
seventhly, repeating the steps until a data set containing the characteristics of the 800-1000 personal body models and walking pace obtained through experimental tests and corresponding assistance amplitude values can be established.
The human body model characteristics of the tester specifically refer to: mastering the information of the age, height, weight, thigh length, shank length, knee diameter, foot length, foot width, ankle bone height and ankle bone width of a tester, wherein the human body characteristics and the unit relationship thereof are shown in table 1, and establishing a sample data set (X, Y) on the basis of the human body characteristics and the unit relationship; wherein X is the walking pace containing the human body model characteristics and obtained by experimental tests, and Y is the corresponding assistance amplitude;
the method for establishing the sample data set (X, Y) comprises the following steps: respectively placing the collected human body model characteristics and walking pace V obtained by experimental tests in each row of an Excel table, and placing the corresponding assistance amplitude A in another row; for example: putting the age in a first row, the height in a second row, the weight in a third row, the thigh length in a fourth row, the shank length in a fifth row, the knee diameter in a sixth row, the foot length in a seventh row, the foot width in an eighth row, the ankle height in a ninth row, the ankle width in a tenth row, the walking pace V in a eleventh row, and the corresponding power-assisted amplitude A in a twelfth row; establishing data sets according to different age groups, establishing a data set for the age group of 20-30 years, establishing a data set for the age group of 31-40 years, establishing a data set for the age group of 41-50 years, establishing a data set for the age group of 51-60 years, establishing a data set for the age group of 61-70 years, establishing a data set for the age group of 71-80 years, and establishing data sets by age groups, so that the predicted boosting amplitude is better close to the actually required boosting amplitude.
(2) After a data set is established and trained, the distribution of a prediction function is directly defined by using a Gaussian process regression mode, wherein the Gaussian process regression is composed of a mean function and a covariance function, and the mathematical expression of the Gaussian process regression is as follows:
f|X～N(m(x),K(X,X)) (1)
wherein X is { X ═ X1,x2,......xnThe walking pace obtained by human body model characteristics and experimental tests; m (x) ═ E [ f (x)]In practical applications, m (x) is often taken as 0, and y is the observed value of the sample; k (X, X) is a covariance matrix; further, an observed value y and a predicted value f can be obtained*The joint distribution of (c) is as follows:
Figure BDA0002810078000000101
in the formula, K (X, X)*)＝K(x*,X)TIs a test point x*I.e. the covariance of any one of the X data with the training set X, K (X)*,x*) Is a test point x*(ii) its own covariance;
Figure BDA0002810078000000102
is the covariance of the sequence of error values, InIs a unit vector of f*The boosting amplitude is predicted according to different human body model characteristics and walking pace;
(3) the form of the test point, i.e. the predicted value, can be obtained according to the properties of the multivariate normal distribution and the conditional distribution, as shown in formula (3):
Figure BDA0002810078000000103
(4) giving the motor power amplitude value which is a predicted power amplitude value obtained by training the data set and optimizing the data set through Gaussian process regression analysis, described in the step (2), so that a tester wears the exoskeleton equipment again to test on the treadmill at different speeds, wherein the test speed is usually selected from three ranges of 6-8Km/h, 8-10Km/h and 4-6Km/h, for example: firstly, testing the comfort level of the assistance amplitude at the speed of 7.5Km/h, then testing at the speed of 8.8Km/h, and finally testing at the speed of 5.5Km/h, wherein if a tester needs to give force to lift and put down the legs freely and conveniently during testing, the current assistance amplitude is not the optimal assistance amplitude, and the assistant rehabilitation effect on bones cannot be achieved; at the moment, the assistance amplitude A in the Excel data set is continuously modified according to the feedback of the current assistance comfort level of a tester, so that the data set is more complete and has stronger universality, the improved data set is retrained and optimized through Gaussian process regression analysis, and the aim is to enable the predicted assistance amplitude to be closer to the actually required assistance amplitude; for example: the assistance amplitude A in the original data set is 30N, but at the moment, the tester feels satisfied, and if the comfortable assistance amplitude A which scores more than 8 points of satisfaction is 35N, the 30N is modified to 35N.
(5) And (3) applying the power-assisted amplitude value after the perfect optimization in the step (4) to a motor, and if a tester feels that the leg is lifted and put down, the tester does not receive the action of self given force any more but completely drives the flexible nylon belt to contract and stretch the given force, namely: the leg is lifted and put down and is not restrained by the action of artificial force any more, the predicted assistance amplitude reaches the expected value, and the assistance amplitude A in the Excel data set is not modified at the moment; only when the leg can be lifted up and put down freely, the injured leg can be helped to walk and jogging under the assistance of the exoskeleton, so that the bone, muscle and joints in the injured leg can be effectively helped to obtain a better recovery effect, and the effect of assisting rehabilitation of the exoskeleton is reflected.
Fig. 2 is a schematic diagram of a device for stretching and contracting a flexible nylon belt by controlling a motor in the optimal power prediction method of the flexible exoskeleton based on gaussian process regression. The black rectangular strips are flexible nylon strips 7 and the aluminum base plate 2 and the spool 4 are tapped with M3 holes 6. The device is a set of detachable device, when needing to give injured or inconvenient shank helping hand of walking, can dress the device on one's body, after muscle and skeleton obtain effective recovery in the injured leg, can dismantle the device. The rectangular boss 1 and the boss 5 are designed on the aluminum base plate 2, the rectangular boss 1 and the boss 5 are both perpendicular to the aluminum base plate, and two round holes are symmetrically formed in the boss 1 on the left side and are used for connecting two gears with the motor 3 and the winding drum 4 respectively to play a role in transmission; four M3 screw holes 6 are formed at four corners of the aluminum base plate 2 for fixing the aluminum base plate 2 to the webbing belt. Two M3 screw holes 6 of symmetry tapping on the reel for fixed flexible nylon area 7, design rectangle boss at the reel both ends moreover, be used for playing the winding effect of direction to flexible nylon area, with reduce unnecessary friction, increase driven stationarity. The fixing devices 8 and 9 are made of 3D printed PLA (poly lactide) materials, square grooves are respectively formed in the contact surfaces of the upper fixing device and the lower fixing device and the connecting piece 10, the connecting piece is fixed, and the fixing devices 8 and 9 are used for connecting the upper flexible nylon belt 7 and the lower flexible nylon belt 7 and carrying the connecting piece. The purpose of the connecting piece is to connect the upper and lower fixing devices 8 and 9 and carry a tension sensor 11 on the connecting piece, and the tension sensor is used for detecting the tension of the leg when the leg is lifted up and put down in real time. The motor drives the winding drum 4, the winding drum 4 is wound with a flexible nylon belt 7, and the flexible nylon belt 7 is parallel to the leg and is tightly attached to help lift the leg. The Y-shaped flexible nylon belt 7 can increase the contact area between the flexible nylon belt and the knee pad, in the lifting process of the flexible nylon belt 7, the Y-shaped flexible nylon belt 7 can play a role in buffering collision and exerting force, the lifting is stable, the flexible nylon belt is not deviated, and two forks of the Y-shaped flexible nylon belt 7 are respectively sewn on the left side and the right side of an opening of a knee 12 of the knee pad. In order to realize flexible assistance for the legs, the device is ensured to be tightly attached to the body of a wearer when the device is worn. Threaded holes 6 of M3 are distributed at intervals of 8CM at the upper end and the lower end of the waistband made of the used braid. The wearing personnel firstly dress the waistband in belly position department and adjust the elasticity degree of waistband, make waistband and belly laminating, the purpose provides steady helping hand effectively and guarantees that the waistband can not become flexible and the landing under motor 3 pivoted effect. Then, the wearer fixes the aluminum base plate 2 having the motor 3 and the drum 4 to the position of the waist belt right above the thighs by M3 screws, fixes the aluminum base plate 2, winds the flexible nylon tape 7 around the drum 4, adjusts the length of the flexible nylon tape wound around the drum so that the flexible nylon tape 7 is parallel to the legs and the knee pad is positioned right at the position of the knee 12, and then firmly wears the knee pad at the position of the knee 12. At the moment, the flexible nylon belt is controlled by the motor to realize the completion of the wearing of the stretching and shrinking device. The device only plays a role in driving and conducting force, and the device plays a role in power-assisted prediction and is realized by establishing a data set and a Gaussian process regression method in the technical scheme.
Searching five volunteers with different age groups and incompletely identical human body characteristics, firstly collecting human body model characteristics of the five volunteers, arranging and classifying the collected information, roughly determining which type of data sets the five volunteers belong to respectively, setting estimated assistance amplitude values which are trained according to the data sets and optimized through Gaussian process regression analysis to a motor, driving a gear connected with the motor end to rotate through a gear at the reel end, enabling the flexible nylon belt on the reel to contract and stretch to give different assistance through the rotation of the gear at the reel end, then, the volunteers wear the exoskeleton equipment to walk on the treadmill at the speed of 7.5Km/h, then walk at the slower speed of 5.5Km/h, and finally walk at the speed of 8.8Km/h, respectively test the comfort degree of the prediction assistance amplitude, and revise the assistance amplitude A in the data set according to the comfort degree value feedback of the volunteers.
And training the improved new data set, and further optimizing the boosting amplitude value of the trained new data set according to a Gaussian process regression analysis prediction method. Let the volunteer carry out the helping hand amplitude comfort level test after optimizing once more, if the volunteer feels the leg and no longer receives self application of force when lifting up and putting down, but the gear on the reel is driven by the gear on the motor completely, thereby make the reel drive flexible nylon belt give required power, the leg lifts up and puts down lightly freely promptly, no longer too much consumption people's physical power, consider that the prediction helping hand amplitude is close to actual required helping hand amplitude this moment, embodied and established the data set and adopted the helping hand amplitude after the gaussian process regression analysis infinitely to be close to required real helping hand amplitude, the helping hand amplitude of having guaranteed the output is more reasonable effective. The leg is light and free to lift and put down, so that the injured leg can still normally walk and move, the problems of a series of bones and joints caused by muscular atrophy and inconsistent length of the injured leg and foot under the condition of not walking for a long time are solved, the rehabilitation of the bones, the muscles and the joints is promoted, and the effect that the walking stick, the wheelchair and other rehabilitation facilities cannot be achieved when the exoskeleton is used for treating and rehabilitating the bones is reflected.

Claims (6)

1. A method for predicting optimal assistance of a flexible exoskeleton based on Gaussian process regression is characterized by comprising the following steps:
(1) establishing data sets of comfortable assistance of different crowds and different pace speeds, establishing human body model characteristics of testers, and establishing data sets (X, Y) of samples on the basis;
(2) after a data set is established and trained, the distribution of a prediction function is directly defined by using a Gaussian process regression mode, wherein the Gaussian process regression is composed of a mean function and a covariance function, and the mathematical expression of the Gaussian process regression is as follows:
f|X～N(m(x),K(X,X)) (1)
wherein X is { X ═ X1,x2,......xnThe walking pace obtained by human body model characteristics and experimental tests; m (x) ═ E [ f (x)]In practical applications, m (x) is often taken as 0, and y is the observed value of the sample; k (X, X) is a covariance matrix; further, an observed value y and a predicted value f can be obtained*The joint distribution of (c) is as follows:
Figure FDA0002810077990000011
in the formula, K (X, X)*)＝K(x*,X)TIs a test point x*I.e. the covariance of any one of the X data with the training set X, K (X)*,x*) Is a test point x*(ii) its own covariance;
Figure FDA0002810077990000012
is the covariance of the sequence of error values, InIs a unit vector of f*The boosting amplitude is predicted according to different human body model characteristics and walking pace;
(3) the form of the test point, i.e. the predicted value, can be obtained according to the properties of the multivariate normal distribution and the conditional distribution, as shown in formula (3):
Figure FDA0002810077990000013
(4) giving a power-assisted amplitude of the motor, wherein the power-assisted amplitude is a predicted power-assisted amplitude obtained by training a data set and optimizing the data set through Gaussian process regression analysis, which is described in the step (2), so that a tester can wear the exoskeleton equipment again to perform tests at different speeds on the treadmill for at least two times, and if the tester needs to give force to lift and put down the legs conveniently and freely during the tests, the current power-assisted amplitude is not the optimal power-assisted amplitude, and the situation that the auxiliary rehabilitation effect on the skeleton cannot be achieved is indicated; at the moment, the assistance amplitude A in the Excel data set is continuously modified according to the feedback of the current assistance comfort level of a tester, so that the data set is more complete and has stronger universality, the improved data set is retrained and optimized through Gaussian process regression analysis, and the aim is to enable the predicted assistance amplitude to be closer to the actually required assistance amplitude;
(5) and (3) applying the power-assisted amplitude value after the perfect optimization in the step (4) to a motor, and if a tester feels that the leg is lifted and put down, the tester does not receive the action of self given force any more but completely drives the nylon belt to contract and stretch the given force, namely: the leg is lifted and put down and is not restrained by the action of artificial force any more, the predicted assistance amplitude reaches the expected value, and the assistance amplitude A in the Excel data set is not modified at the moment; only when the leg can be lifted up and put down freely, the injured leg can be helped to walk and jogging under the assistance of the exoskeleton, so that the bone, muscle and joints in the injured leg can be effectively helped to obtain a better recovery effect, and the effect of assisting rehabilitation of the exoskeleton is reflected.
2. The optimal assistance prediction method for the flexible exoskeleton of claim 1 based on the Gaussian process regression is characterized in that the method for establishing the data set of comfortable assistance of different people and different walking speeds in the step (1) comprises the following steps:
firstly, leading testers with different age groups and different human body characteristics to wear exoskeleton equipment to walk on a running machine at the speed of 7.5 Km/h;
firstly, setting an initial boosting amplitude of 24N by a motor carried on the exoskeleton through driving the flexible nylon belt to contract and stretch;
thirdly, the comfort level of the testing personnel is scored, namely: the tester scores the comfort degree of the leg, the comfort degree score range is 0-10, and the higher the score, the more suitable the amplitude of the assistance is;
if the tester feels that the output boosting amplitude is not comfortable, namely, the evaluation time is less than 5 minutes, the rotating speed of the motor is adjusted to drive the flexible nylon belt to contract and stretch, so that the boosting amplitudes with different sizes are output, the amplitude of comfortable boosting is in a more reasonable and effective range, the leg pressure is effectively relieved, and the bones, muscles and joints are helped to recover;
if the tester is satisfied with the current output power-assisted amplitude and marks a satisfied time division of more than 8 minutes, recording the current power-assisted amplitude as A and the walking pace as V;
sixthly, changing the current walking pace V by the tester, and repeating the step two to the step five;
seventhly, repeating the steps until a data set containing human body model characteristics, walking pace obtained through experimental tests and corresponding assistance amplitude values can be established.
3. The method of claim 2, wherein the number of the human model features in step (c) is not less than 100.
4. The optimal assistance prediction method for the flexible exoskeleton based on the Gaussian process regression as claimed in claim 1, wherein the step (1) of establishing the human body model features of the test person specifically comprises the steps of: the information of the age, height, weight, thigh length, shank length, knee diameter, foot length, foot width, ankle height and ankle width of the test person is obtained, and the human body characteristics and the unit relationship thereof are shown in table 1:
TABLE 1 human body characteristics and unit relationship corresponding table
Figure FDA0002810077990000031
Figure FDA0002810077990000041
Establishing a sample data set (X, Y) on the basis of the data set; wherein X is the walking pace containing the human body model characteristics and obtained by experimental tests, and Y is the corresponding assistance amplitude.
5. The optimal assistance prediction method for the flexible exoskeleton based on Gaussian process regression as claimed in claim 4, wherein the method for establishing the sample data set (X, Y) in the step (1) is as follows: respectively placing the collected human body model characteristics and walking pace V obtained by experimental tests in each row of an Excel table, and placing the corresponding assistance amplitude A in another row; and data sets are respectively established according to age groups.
6. The optimal assistance prediction method for the flexible exoskeleton based on the Gaussian process regression as claimed in claim 5, wherein the step of respectively establishing the data sets according to the age groups specifically comprises the steps of: establishing a data set for 20-30 years old, establishing a data set for 31-40 years old, establishing a data set for 41-50 years old, establishing a data set for 51-60 years old, establishing a data set for 61-70 years old, establishing a data set for 71-80 years old, and establishing data sets by different ages.
CN202011387447.0A 2020-12-01 2020-12-01 Optimal assistance prediction method for flexible exoskeleton based on Gaussian process regression Pending CN112733422A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011387447.0A CN112733422A (en) 2020-12-01 2020-12-01 Optimal assistance prediction method for flexible exoskeleton based on Gaussian process regression

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011387447.0A CN112733422A (en) 2020-12-01 2020-12-01 Optimal assistance prediction method for flexible exoskeleton based on Gaussian process regression

Publications (1)

Publication Number Publication Date
CN112733422A true CN112733422A (en) 2021-04-30

Family

ID=75598142

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011387447.0A Pending CN112733422A (en) 2020-12-01 2020-12-01 Optimal assistance prediction method for flexible exoskeleton based on Gaussian process regression

Country Status (1)

Country Link
CN (1) CN112733422A (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160331560A1 (en) * 2015-05-11 2016-11-17 The Hong Kong Polytechnic University Interactive Exoskeleton Robotic Knee System
CN109543715A (en) * 2018-10-23 2019-03-29 武汉理工大学 A kind of ship air route is extracted and the method for track deviation detection
CN110110633A (en) * 2019-04-28 2019-08-09 华东交通大学 Method for automatically identifying and analyzing hemiplegic gait based on machine learning
CN110710976A (en) * 2019-10-14 2020-01-21 深圳市迈步机器人科技有限公司 Storage medium for personalized gait generation method, control system and exoskeleton robot

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160331560A1 (en) * 2015-05-11 2016-11-17 The Hong Kong Polytechnic University Interactive Exoskeleton Robotic Knee System
CN109543715A (en) * 2018-10-23 2019-03-29 武汉理工大学 A kind of ship air route is extracted and the method for track deviation detection
CN110110633A (en) * 2019-04-28 2019-08-09 华东交通大学 Method for automatically identifying and analyzing hemiplegic gait based on machine learning
CN110710976A (en) * 2019-10-14 2020-01-21 深圳市迈步机器人科技有限公司 Storage medium for personalized gait generation method, control system and exoskeleton robot

Similar Documents

Publication Publication Date Title
Huo et al. Lower limb wearable robots for assistance and rehabilitation: A state of the art
Gasparri et al. Proportional joint-moment control for instantaneously adaptive ankle exoskeleton assistance
CN105873555A (en) Joint movement assistance device
WO2008124025A1 (en) Powered orthosis
US20100312152A1 (en) Smart gait rehabilitation system for automated diagnosis and therapy of neurologic impairment
Khomami et al. A survey on soft lower limb cable-driven wearable robots without rigid links and joints
CN104869969A (en) Soft exosuit for assistance with human motion
CN105411811A (en) Modularized multifunctional rehabilitation training device
CN102665827A (en) Improved rehabilitation and exercise machine
CN109830158B (en) Dynamic knee joint simulation device
Sridar et al. Evaluating immediate benefits of assisting knee extension with a soft inflatable exosuit
Chen et al. A pediatric knee exoskeleton with real-time adaptive control for overground walking in ambulatory individuals with cerebral palsy
Zhong et al. Fuzzy logic compliance adaptation for an assist-as-needed controller on the Gait Rehabilitation Exoskeleton (GAREX)
Conway et al. The functional utilization of propulsive capacity during human walking
CN113599781A (en) Wearable walking fitness system based on exoskeleton, control method and storage medium
JP6858400B2 (en) Walking training device, walking diagnostic device, weight unloading device, and walking diagnostic method
CN112733422A (en) Optimal assistance prediction method for flexible exoskeleton based on Gaussian process regression
CN113539419A (en) Method and system for evaluating stroke gymnastics effect
JP2004167056A (en) Gait training apparatus
Fang et al. Bilateral vs. Paretic-Limb-Only ankle exoskeleton assistance for improving hemiparetic gait: A case series
Koenig et al. Voluntary gait speed adaptation for robot-assisted treadmill training
CN211273391U (en) Utilize recovered fitness equipment of upper limbs pulling force supplementary low limbs exercise
CN106074076A (en) The medical rehabilitation robot system of multi-modal perceptible feedback
Chen et al. Biomechanical and physiological evaluation of biologically-inspired hip assistance with belt-type soft exosuits
Wang et al. Development of an automated assistive trainer inspired by neuro-developmental treatment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20210430