CN109009586A - A kind of myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint - Google Patents

A kind of myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint Download PDF

Info

Publication number
CN109009586A
CN109009586A CN201810664049.5A CN201810664049A CN109009586A CN 109009586 A CN109009586 A CN 109009586A CN 201810664049 A CN201810664049 A CN 201810664049A CN 109009586 A CN109009586 A CN 109009586A
Authority
CN
China
Prior art keywords
wrist
angle
joint
wrist joint
neural network
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.)
Granted
Application number
CN201810664049.5A
Other languages
Chinese (zh)
Other versions
CN109009586B (en
Inventor
张小栋
孙晓峰
陆竹风
李瀚哲
李睿
郭健
杨昆才
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.)
Xian Jiaotong University
Original Assignee
Xian Jiaotong University
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 Xian Jiaotong University filed Critical Xian Jiaotong University
Priority to CN201810664049.5A priority Critical patent/CN109009586B/en
Publication of CN109009586A publication Critical patent/CN109009586A/en
Application granted granted Critical
Publication of CN109009586B publication Critical patent/CN109009586B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61F—FILTERS IMPLANTABLE INTO BLOOD VESSELS; PROSTHESES; DEVICES PROVIDING PATENCY TO, OR PREVENTING COLLAPSING OF, TUBULAR STRUCTURES OF THE BODY, e.g. STENTS; ORTHOPAEDIC, NURSING OR CONTRACEPTIVE DEVICES; FOMENTATION; TREATMENT OR PROTECTION OF EYES OR EARS; BANDAGES, DRESSINGS OR ABSORBENT PADS; FIRST-AID KITS
    • A61F2/00—Filters implantable into blood vessels; Prostheses, i.e. artificial substitutes or replacements for parts of the body; Appliances for connecting them with the body; Devices providing patency to, or preventing collapsing of, tubular structures of the body, e.g. stents
    • A61F2/50—Prostheses not implantable in the body
    • A61F2/68—Operating or control means
    • A61F2/70—Operating or control means electrical
    • A61F2/72—Bioelectric control, e.g. myoelectric
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/389—Electromyography [EMG]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Public Health (AREA)
  • Biomedical Technology (AREA)
  • Veterinary Medicine (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Animal Behavior & Ethology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Artificial Intelligence (AREA)
  • Surgery (AREA)
  • Physiology (AREA)
  • Cardiology (AREA)
  • Psychiatry (AREA)
  • Signal Processing (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Fuzzy Systems (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Transplantation (AREA)
  • Vascular Medicine (AREA)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
  • Prostheses (AREA)

Abstract

The invention discloses a kind of myoelectricity continuous decoding methods of the man-machine natural driving angle of artificial hand wrist joint, this method allows disabled person to imagine that it lacks hand and perfects the wrist joints sporting that hand does equal angular, the surface electromyogram signal of acquisition disabled person's deformed limb forearm and strong side wrist joints sporting angle at this time, to establish the man-machine natural driving angle model of surface electromyogram signal continuous decoding artificial hand wrist joint.Operation of the present invention is simple and accuracy is high.Three-dimensional motion, which captures system, can be realized simultaneously high-resolution and high capture frequency, and the joint angles thus calculated have high precision；Compared with traditional Worn type angular transducer, compression type will not be caused to interfere surface electromyogram signal；With the interface with myoelectric apparatus synchronous acquisition, it can be realized three-dimensional motion capture angle and carried out simultaneously with surface electromyogram signal acquisition.In addition, the sample frequency of myoelectric apparatus is 2048Hz, it can real-time collection surface electromyography signal situation of change.

Description

A kind of myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint
Technical field
The invention belongs to Intelligent artificial hands and raw electromechanical integration technology area, are related to a kind of man-machine drive naturally of artificial hand wrist joint The myoelectricity continuous decoding method of dynamic angle.
Background technique
It is shown according to China Disabled Persons' Federation's updated statistics, physical disabilities number is about 24,720,000 people now in China, wherein hand Portion's disabled patient reaches up to ten million.The missing of hand function not only influences the life and work of disabled person, and it is heavy more to bring to its psychology It beats again and hits.Tradition artificial hand only serves rhetorical function, is not able to satisfy needed for its daily life.The appearance of Intelligent artificial hand compensates for tradition Do evil through another person function deficiency, wherein myoelectrically controlled hand because easy to wear, precise control, it is powerful due to have become a hot topic of research.
Human body surface myoelectric signal is a kind of bioelectrical signals, can objectively react the motion state of human body and can surpass Preceding actual act generates, and has foresight, and the perception of human motion intention may be implemented.Existing EMG-controlling prosthetic hand, major part will Research emphasis is placed on using the surface electromyogram signal identification manpower classification of motion to realize that artificial hand grasping movement is predicted, however in vacation During hands movement, carpal driving angle has been largely fixed the flexibility of artificial hand operation, so in order to realize vacation Hand preferably personalizes, and is seemed pole using the man-machine natural driving angle of surface electromyogram signal decoding artificial hand wrist joint of arm remaining Its is important.Based on this, suitable artificial hand is provided using arm surface electromyography signal continuous decoding wrist joint movement angle The man-machine natural driving angle of wrist joint is the key point of current research.
Summary of the invention
It is an object of the invention to overcome the above-mentioned prior art, provide a kind of artificial hand wrist joint man-machine driving naturally The myoelectricity continuous decoding method of angle, this method allow disabled person to imagine that it lacks hand and perfects the wrist joint fortune that hand does equal angular It moves, at this time the surface electromyogram signal of acquisition disabled person's deformed limb forearm and strong side wrist joints sporting angle, to establish surface myoelectric The man-machine natural driving angle model of signal continuous decoding artificial hand wrist joint.It realizes and utilizes deformed limb forearm surface electromyogram signal continuous solution The man-machine natural driving angle of code artificial hand wrist joint, to meet needed for hand disability patient's daily life.
In order to achieve the above objectives, the present invention is achieved by the following scheme:
A kind of method of the man-machine natural driving angle of human arm surface electromyogram signal continuous decoding artificial hand wrist joint, including under State step:
Step 1: recording healthy side hand wrist joints sporting three-dimensional coordinate using motion capture system, transported by manpower wrist joint Dynamic modeling method of learning calculates wrist flex/stretching angle.
In the above method, chooses reasonable wrist joints sporting and capture scheme and establish wrist joint local coordinate system, establish wrist Joint motions model calculates angle change of the wrist joint in bending stretching process using kinematic method.Finally obtain wrist Arthrogryposis stretching angle formula is as follows:
θ=arccos (Ti2·i1)
In formula, T is transition matrix of the wrist joint local coordinate system to elbow joint local coordinate system, i1、i2Respectively elbow joint With the unit vector of wrist joint sagittal axis direction.
Step 2: utilizing the surface electromyogram signal of the six pieces of muscle in myoelectricity Acquisition Instrument synchronous acquisition forearm remaining side.Described six Block remaining muscle is respectively musculus extensor carpi radialis longus, musculus flexor carpi radialis, ulnar side wrist extensor hallucis longus, musculus flexor carpi ulnaris, musculus extensor digitorum, refers to shallowly Musculus flexor.
Step 3: carrying out pretreatment and feature extraction to collected six channel surface electromyogram signal.Due to surface myoelectric Instrument sample frequency be 2048Hz, three-dimensional motion capture system sampling frequency be 100Hz, so to surface electromyogram signal characteristic value into Surface electromyogram signal and wrist joints sporting data sample frequency having the same are realized in row resampling.
Step 4: establishing BP neural network using the method for machine learning, human arm surface electromyogram signal continuous solution is realized The angle of code wrist joint flexion/extension.The network parameter of BP neural network is set first, secondly BP neural network is instructed Practice, finally it is tested.
In the above method, three layers of BP neural network are constructed, the myoelectricity for extracting surface electromyogram signal enlivens characteristic strength value work For network inputs, being exported by the joint angles that wrist joints sporting model calculates as network, 10 neurons are arranged in middle layer, Each neuron uses Sigmoid action function.
Step 5: acquiring residual side forearm surface electromyogram signal, collected surface electromyogram signal is subjected to pretreatment and spy Sign is extracted.Myoelectricity is enlivened into characteristic strength value input wrist joint angle continuous decoding model, the wrist for exporting consecutive variations is closed Movement angle is saved, and calculates the linearly dependent coefficient between the joint angles of neural network forecast joint angles and kinematic calculation out, Judge the accuracy of the man-machine natural driving angle of human arm surface electromyogram signal continuous decoding artificial hand wrist joint.
Compared with prior art, the invention has the following advantages:
Operation of the present invention is simple and accuracy is high.Three-dimensional motion, which captures system, can be realized simultaneously high-resolution and high capture Frequency, the joint angles thus calculated have high precision；It, will not be to surface myoelectric compared with traditional Worn type angular transducer Signal causes compression type to interfere；With the interface with myoelectric apparatus synchronous acquisition, it can be realized three-dimensional motion and capture angle and surface Electromyographic signal collection carries out simultaneously.In addition, the sample frequency of myoelectric apparatus is 2048Hz, collection surface electromyography signal can become in real time Change situation.
The present invention establishes the prediction model of surface electromyogram signal continuous decoding artificial hand wrist joint driving angle.Skeletal muscle It stretches and shrinks drive wrist joints sporting, in muscle contraction, corresponding surface electromyogram signal amplitude has difference Variation, it is possible to using the myoelectricity of surface electromyogram signal enliven intensity continuous decoding artificial hand wrist joint it is man-machine naturally driving angle Degree.It allows disabled person to imagine that lack hand synchronous with the wrist joint for perfecting hand and carries out identical movement, by Mental imagery training Afterwards, the man-machine natural driving angle of wrist joint done evil through another person using side wrist joints sporting angle is perfected as residual side.Mode input is residual Side forearm surface electromyogram signal avoids and models bring individual sex differernce to disabled person by abled person's modeling, to realize The man-machine natural driving angle of surface electromyogram signal continuous decoding artificial hand wrist joint, has in fields such as biologic medical, human engineerings Potential application value.
Detailed description of the invention
Fig. 1 is the method block diagram of the man-machine natural driving angle of arm surface electromyography signal continuous decoding artificial hand wrist joint；
Fig. 2 is right side upper limb mark point position and coordinate system setting schematic diagram；Wherein P1For elbow joint, P2For forearm oar At side, P3For wrist outside, P4At on the inside of wrist, P5For right hand middle finger metacarpophalangeal joint；
Fig. 3 is wrist joint bending stretching angle schematic diagram of calculation result；
Fig. 4 is that six channel surface electromyogram signal myoelectricity of arm enlivens characteristic strength value；
Fig. 5 is BP neural network algorithm basic flow chart；
Fig. 6 is the man-machine natural driving angle result figure of arm surface electromyography signal continuous decoding artificial hand wrist joint.
Specific embodiment
The invention will be described in further detail with reference to the accompanying drawing:
Referring to Fig. 1, the present invention captures system record healthy side hand wrist joint using three-dimensional motion and is bent stretching process kinematics Data calculate the angle of wrist joint bending stretching, extension；The surface electromyogram signal of the residual side forearm muscle of myoelectric apparatus synchronous acquisition, through pre- Processing and feature extraction obtain myoelectricity and enliven strength characteristic；Myoelectricity is enlivened into input of the strength characteristic as BP neural network, wrist Output of the arthrogryposis stretching angle as BP neural network, is trained BP neural network, sets error range, meets and misses Stop iteration after poor condition, obtains the stable man-machine natural driving angle model of arm decoding artificial hand wrist joint；Recently enter survey The residual side surface electromyography signal of examination enlivens intensity, exports the man-machine natural driving angle of artificial hand wrist joint of prediction.Specific embodiment party Case is as follows:
Step 1: calculating wrist joint by kinematic data is bent stretching angle, the specific method is as follows:
(1) the motion-captured mark point mark position of wrist joint is chosen and local joint establishment of coordinate system:
Fig. 2 gives the stickup scheme of right side upper limb mark point, and in order to guarantee that mark point is not blocked, subject takes seat Appearance state, right side large arm and forearm level are lifted and are remain stationary, and the back of the hand does wrist joint bending stretching, extension upwards.According to kinematics original It manages, there are 6 freedom degrees in each 3 dimension space of rigid body, to determine the pose of motion rigid body in three dimensions, need to know just The position coordinates of 3 points of non-colinear on body.So a mark point is arranged in elbow joint, it is denoted as P1, Radial Forearm setting one A mark point, is denoted as P2, on the outside of wrist and a mark point is respectively set in inside, is denoted as P3And P4, then refer to metacarpophalangeal in the right hand A mark point is arranged in joint, is denoted as P5。
Elbow joint local coordinate system is as shown in Figure 2:
Elbow joint local coordinate origin is P1, P1With P2Line is x-axis, and P is directed toward in direction2；By P1、P2、P33 points of compositions are flat The normal in face is y-axis, and direction is directed toward on the inside of body；By right hand rule it is found that the normal vector that x-axis and y-axis constitute plane is z-axis, Direction is upward.The unit vector calculation formula of x, y, z axis is respectively as follows:
k1=i1×j1
Wrist joint local coordinate system is as shown in Figure 2:
Wrist joint local coordinate origin is P3With P4Line midpoint, origin and P5Line is x-axis, and P is directed toward in direction5；By P3、 P4、P5The normal vector of 3 points of composition planes is z-axis, and direction is downward；By right hand rule it is found that x-axis and z-axis constitute the normal direction of plane Amount is y-axis, and direction is directed toward on the inside of body.The unit vector calculation formula of x, y, z axis are as follows:
k2=i2×j2
(2) wrist joint movement angle calculates:
Flexion/extension angle of the wrist joint in human body sagittal plane is carried out on the basis of elbow joint and wrist joint local coordinate system Solution, the flexion/extension angle of wrist is indicated with θ.
θ=arccos (Ti2·i1)
In formula, T is transition matrix of the wrist joint local coordinate system to elbow joint local coordinate system, i1、i2Respectively elbow joint With the unit vector of wrist joint sagittal axis direction.
Fig. 3 gives the calculated wrist joint flexion/extension angle change of thus method.It defines when palm hand prosposition is set and is 0 degree, bending angle is positive, stretching angle is negative, and wrist joint bending stretching routine range is -60 °~75 °, it can be seen that Calculated wrist flex/the stretching angle of the method is rationally and in manpower flexion/extension motion range.
Step 2: utilizing the surface electromyogram signal of the related six pieces of muscle in the residual side of myoelectric apparatus synchronous acquisition forearm.Specific method It is as follows:
The discovery of manpower forearm muscle is analyzed, musculus extensor carpi radialis longus, ulnar side wrist extensor hallucis longus and musculus extensor digitorum and wrist stretching, extension are dynamic Make close relation, musculus flexor carpi radialis, musculus flexor carpi ulnaris and musculus flexor digitorum sublimis and wrist joint flexure operation close relation, thus choose with Upper six pieces of muscle is as surface electromyogram signal signal source.It is carried out disinfection first with alcohol to subject's forearm, reduces skin surface Interference of the grease to surface electromyogram signal；Secondly differential type electrode is pasted onto six pieces of muscle tables along meat fiber direction respectively Face, electromyography signal reset is in elbow joint；Finally realize that myoelectric apparatus is synchronous with three-dimensional motion capture system using trigger Acquisition.
Step 3: collected six channels arm surface electromyography signal is pre-processed, extracts myoelectricity and enliven intensity spy Sign finally enlivens intensity to arm surface electromyogram signal and carries out resampling, and uniform surface electromyography signal is adopted with wrist joint angle value Sample frequency.The specific method is as follows:
(1) arm surface electromyography signal pre-processes:
Collected arm surface electromyography signal would generally be mixed with noise and interference, including equipment intrinsic noise, surrounding are made an uproar Acoustic jamming, 50Hz Hz noise and artefact noise etc..Surface electromyogram signal effective frequency range is 20-500Hz, in order to reduce The interference of noise signal carries out 20-500Hz bandpass filtering to surface electromyogram signal using 4 rank Butterworth filters, recycles Notch filter removes 50Hz Hz noise.
(2) arm surface electromyography signal feature extraction:
In order to realize the man-machine natural driving angle of arm surface electromyography signal continuous decoding artificial hand wrist joint, need to find hand Arm surface electromyogram signal intensity and man-machine driven between joint angles naturally of artificial hand contact, and can extract the flesh of surface electromyogram signal Electricity enlivens strength characteristic.Using envelope method, all-wave arrangement is carried out to the surface electromyogram signal after filtering, then carry out low pass filtered Wave selects cutoff frequency for 4~10Hz, obtains surface electromyogram signal peak change, as shown in Figure 4.
Step 4: the method for taking machine learning, establishes BP neural network, arm surface electromyography signal continuous decoding is realized The man-machine natural driving angle of wrist joint of doing evil through another person.The specific method is as follows:
(1) building of BP neural network:
Under normal conditions, three layers of BP neural network is able to solve most pattern recognition problems, so constructing three layers here BP neural network.Input layer is surface myoelectric characteristic strength value, takes b=6 here, i.e. input layer has 6 neuron nodes；It is defeated Layer is wrist flex stretching angle out, i.e. output layer has 1 neuron node；Middle layer neuron node number is by following experience Publicity determines:
Wherein, g is middle layer node number, and b is input layer number, and c is output layer node number, and d=1~10 are Regulating constant.Here d=8, i.e. g=10 are taken.
(2) training of BP neural network:
In Fig. 5, if network inputs vector is X=[x1 x2 x3 x4 x5 x6]T, network output vector is Y=[y]T, intermediate The neuron of layer network exports are as follows:
Output layer output are as follows:
Wherein, neuron operation function are as follows:
Define error function:
Wherein,
N is sample number, and m is sample points in each sample, dpiTo be stretched by the calculated wrist joint bending of kinematic data Open up angle, ypiWrist joint to be estimated by BP neural network is bent stretching angle, and q is the network number of plies.
The local minimum of E is found using gradient descent method, each connection weight is both needed to along E to connection weight derivative Opposite direction amendment.If error function in the ideal range, stops iteration, otherwise continue to be modified connection weight until accidentally Difference is sufficiently small.
(3) test of BP neural network:
The BP neural network that 3 groups of arm surface electromyography signal characteristic values are completed to training is inputted, then it is curved to export 3 groups of wrists Bent stretching angle predicted value.It calculates the wrist flex stretching angle of BP neural network input and is really captured through three-dimensional motion and be The related coefficient between angle that the kinematic data that system captures resolves, to react linearly related degree between the two.Phase It is as follows to close coefficient formulas:
Wherein, Cov (X, Y) is the covariance of X and Y, and Var [X] and Var [Y] are respectively the variance of X and Y.Related coefficient | ρxy|≤1, | ρxy| indicate that X and Y degree of correlation is higher closer to 1, | ρxy| indicate that X and Y degree of correlation is lower closer to 0.
It extracts arm surface electromyography signal myoelectricity and enlivens strength characteristic, artificial hand wrist joint is predicted using BP neural network method The decoding result of man-machine nature driving angle is as shown in Figure 6.As seen from the figure, the method, which can be stablized, effectively realizes arm surface Electromyography signal continuous decoding wrist joint is bent stretching angle, can be used for the man-machine natural drive control of myoelectricity Intelligent artificial hand.
The present invention is a kind of method of man-machine natural driving angle of arm surface electromyography signal continuous decoding artificial hand wrist joint, To realize through forearm surface electromyogram signal amplitude variation identification artificial hand wrist joint bending stretching, extension driving angle, increase Intelligent artificial hand Personalize nature operation.This method is extracted musculus extensor carpi radialis longus, musculus flexor carpi radialis, ulnar side wrist extensor hallucis longus, musculus flexor carpi ulnaris, is referred to The total surface electromyogram signal of six pieces of extensor, musculus flexor digitorum sublimis muscle in wrist flex stretching process, is bent in conjunction with wrist joint and stretches Angle change value in the process establishes Nonlinear Mapping model using BP neural network, reaches arm surface electromyography signal continuous solution The man-machine natural driving angle of code artificial hand wrist joint, flexibly controls the purpose of artificial hand wrist action.The invention model built is reliable, Angle predicts that accuracy is high, preferably uses Intelligent artificial hand to carry out nature operation conducive to hand disability patient, meet its life and Artificial hand functional requirement in work has good Social benefit and economic benefit.
The above content is merely illustrative of the invention's technical idea, and this does not limit the scope of protection of the present invention, all to press According to technical idea proposed by the present invention, any changes made on the basis of the technical scheme each falls within claims of the present invention Protection scope within.

Claims (5)

1. a kind of myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint, which is characterized in that including following step It is rapid:
Step 1: recording healthy side hand wrist joints sporting three-dimensional coordinate using motion capture system, pass through wrist joint Kinematic Model Method calculates wrist flex/stretching angle；
Step 2: utilizing the surface electromyogram signal of the six pieces of muscle in myoelectricity Acquisition Instrument synchronous acquisition forearm remaining side；Six pieces of remaining muscle Respectively musculus extensor carpi radialis longus, musculus flexor carpi radialis, ulnar side wrist extensor hallucis longus, musculus flexor carpi ulnaris, musculus extensor digitorum and musculus flexor digitorum sublimis；
Step 3: pretreatment and feature extraction are carried out to collected six channel surface electromyogram signal；To surface electromyogram signal feature Value carries out resampling, realizes surface electromyogram signal and wrist joints sporting data sample frequency having the same；
Step 4: using the method for machine learning, establishing BP neural network, realize that arm surface electromyography signal continuous decoding wrist closes Bent-segment/stretching, extension angle；The network parameter of BP neural network is set first, secondly BP neural network is trained, finally It is tested；
Step 5: acquiring residual side forearm surface electromyogram signal, collected surface electromyogram signal is subjected to pretreatment and is mentioned with feature It takes；Myoelectricity is enlivened into characteristic strength value input wrist joint angle continuous decoding model, exports the wrist joint fortune of consecutive variations Dynamic angle, and calculate the linearly dependent coefficient between the joint angles of neural network forecast joint angles and kinematic calculation out, judgement The accuracy of the man-machine natural driving angle of arm surface electromyography signal continuous decoding artificial hand wrist joint.
2. the myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint according to claim 1, feature It is, step 1 calculates wrist joint bending stretching angle, and the specific method is as follows:
1-1) the motion-captured mark point mark position of wrist joint is chosen and local joint establishment of coordinate system:
One mark point is set in elbow joint, is denoted as P1, Radial Forearm be arranged a mark point, be denoted as P2, on the outside of wrist and A mark point is respectively set in inside, is denoted as P3And P4, then refer to metacarpophalangeal joint one mark point of setting in the right hand, it is denoted as P5；
Elbow joint local coordinate origin is P1, P1With P2Line is x-axis, and P is directed toward in direction2；By P1、P2、P3The method of 3 points of composition planes Line is y-axis, and direction is directed toward on the inside of body；Obtained by right hand rule, the normal vector that x-axis and y-axis constitute plane is z-axis, direction to On；The unit vector calculation formula of x, y, z axis is respectively as follows:
k1=i1×j1
Wrist joint local coordinate origin is P3With P4Line midpoint, origin and P5Line is x-axis, and P is directed toward in direction5；By P3、P4、P5Three The normal vector that point constitutes plane is z-axis, and direction is downward；It is obtained by right hand rule, x-axis and z-axis constitute the normal vector of plane as y Axis, direction are directed toward on the inside of body；The unit vector calculation formula of x, y, z axis are as follows:
k2=i2×j2
1-2) wrist joint movement angle calculates:
Elbow joint with asking for flexion/extension angle of the wrist joint in human body sagittal plane is carried out on the basis of wrist joint local coordinate system Solution, the flexion/extension angle of wrist is indicated with θ；
θ=arccos (Ti2·i1)
In formula, T is transition matrix of the wrist joint local coordinate system to elbow joint local coordinate system, i1、i2Respectively elbow joint and wrist The unit vector of joint sagittal axis direction.
3. the myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint according to claim 1, feature It is, the specific method is as follows for step 3:
3-1) arm surface electromyography signal pre-processes:
20-500Hz bandpass filtering is carried out to surface electromyogram signal using 4 rank Butterworth filters, notch filter is recycled to go Except 50Hz Hz noise；
3-2) arm surface electromyography signal feature extraction:
Using envelope method, all-wave arrangement is carried out to the surface electromyogram signal after filtering, then carry out low-pass filtering, selection cut-off Frequency is 4~10Hz, obtains surface electromyogram signal peak change.
4. the myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint according to claim 1, feature It is, in step 4, constructs three layers of BP neural network, the myoelectricity for extracting surface electromyogram signal enlivens characteristic strength value as network Input is exported by the joint angles that wrist joints sporting model calculates as network, and 10 neurons, each mind is arranged in middle layer Sigmoid action function is used through member.
5. the myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint according to claim 1 or 4, special Sign is that the specific method is as follows for step 4:
4-1) the building of BP neural network:
Construct three layers of BP neural network；Input layer is surface myoelectric characteristic strength value, takes b=6, i.e. input layer has 6 neurons Node；Output layer is wrist flex stretching angle, i.e. output layer has 1 neuron node；Middle layer neuron node number by Following experience publicity determines:
Wherein, g is middle layer node number, and b is input layer number, and c is output layer node number, and d=1~10 are to adjust Constant；Take d=8, i.e. g=10；
4-2) the training of BP neural network:
If network inputs vector is X=[x1 x2 x3 x4 x5 x6]T, network output vector is Y=[y]T, the mind of mid-level network It is exported through member are as follows:
Output layer output are as follows:
Wherein, neuron operation function are as follows:
Define error function:
Wherein,
N is sample number, and m is sample points in each sample, dpiTo be bent extension angle by the calculated wrist joint of kinematic data Degree, ypiWrist joint to be estimated by BP neural network is bent stretching angle, and q is the network number of plies；
The local minimum of E is found using gradient descent method, each connection weight is both needed to the negative side along E to connection weight derivative To amendment；If error function in the ideal range, stops iteration, otherwise continue to be modified connection weight until error foot It is enough small；
4-3) the test of BP neural network:
The BP neural network that 3 groups of arm surface electromyography signal characteristic values are completed to training is inputted, then can export 3 groups of wrist flex and stretch Open up angle predicted value；Calculate BP neural network input wrist flex stretching angle with really through three-dimensional motion capture system catch The related coefficient between angle that the kinematic data caught resolves, to react linearly related degree between the two；Phase relation Number calculation formula is as follows:
Wherein, Cov (X, Y) is the covariance of X and Y, and Var [X] and Var [Y] are respectively the variance of X and Y；Related coefficient | ρxy|≤ 1, | ρxy| indicate that X and Y degree of correlation is higher closer to 1, | ρxy| indicate that X and Y degree of correlation is lower closer to 0.
CN201810664049.5A 2018-06-25 2018-06-25 Myoelectric continuous decoding method for man-machine natural driving angle of artificial wrist joint Active CN109009586B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810664049.5A CN109009586B (en) 2018-06-25 2018-06-25 Myoelectric continuous decoding method for man-machine natural driving angle of artificial wrist joint

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810664049.5A CN109009586B (en) 2018-06-25 2018-06-25 Myoelectric continuous decoding method for man-machine natural driving angle of artificial wrist joint

Publications (2)

Publication Number Publication Date
CN109009586A true CN109009586A (en) 2018-12-18
CN109009586B CN109009586B (en) 2020-07-28

Family

ID=64610702

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810664049.5A Active CN109009586B (en) 2018-06-25 2018-06-25 Myoelectric continuous decoding method for man-machine natural driving angle of artificial wrist joint

Country Status (1)

Country Link
CN (1) CN109009586B (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109259739A (en) * 2018-11-16 2019-01-25 西安交通大学 A kind of myoelectricity estimation method of wrist joint motoring torque
CN110801226A (en) * 2019-11-01 2020-02-18 西安交通大学 Human knee joint moment testing system method based on surface electromyographic signals and application
CN110827987A (en) * 2019-11-06 2020-02-21 西安交通大学 Myoelectricity continuous prediction method and system for wrist joint torque in multi-grabbing mode
CN111374808A (en) * 2020-03-05 2020-07-07 北京海益同展信息科技有限公司 Artificial limb control method and device, storage medium and electronic equipment
CN111616848A (en) * 2020-06-02 2020-09-04 中国科学技术大学先进技术研究院 Five-degree-of-freedom upper arm prosthesis control system based on FSM
CN111920416A (en) * 2020-07-13 2020-11-13 张艳 Hand rehabilitation training effect measuring method, storage medium, terminal and system
CN112587242A (en) * 2020-12-11 2021-04-02 天津大学医疗机器人与智能系统研究院 Surgical robot master hand simulation method, master hand and application
WO2022001771A1 (en) * 2020-06-29 2022-01-06 京东科技信息技术有限公司 Artificial limb control method, device and system and storage medium
CN114224577A (en) * 2022-02-24 2022-03-25 深圳市心流科技有限公司 Training method and device for intelligent artificial limb, electronic equipment, intelligent artificial limb and medium
CN117953413A (en) * 2024-03-27 2024-04-30 广东工业大学 Electromyographic signal validity judging method, electromyographic signal validity judging device, electromyographic signal validity judging equipment and storage medium

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2163570A (en) * 1984-08-24 1986-02-26 Hanger & Co Ltd J E Artificial hand
US20130253606A1 (en) * 2012-03-20 2013-09-26 In Chan Youn Peripheral nerve interface system and method for prosthetic hand control
CN104665828A (en) * 2013-11-27 2015-06-03 中国科学院深圳先进技术研究院 System and method based on electromyographic signal controlling remote controller
CN105615890A (en) * 2015-12-24 2016-06-01 西安交通大学 Angle and myoelectricity continuous decoding method for human body lower limb walking joint
CN105963100A (en) * 2016-04-19 2016-09-28 西安交通大学 Patient movement demand-based assistance lower limb rehabilitation robot self-adaptation control method
CN106236336A (en) * 2016-08-15 2016-12-21 中国科学院重庆绿色智能技术研究院 A kind of myoelectric limb gesture and dynamics control method
CN106923942A (en) * 2017-02-15 2017-07-07 上海术理智能科技有限公司 Upper and lower extremities motion assistant system based on the control of human body electromyographic signal

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2163570A (en) * 1984-08-24 1986-02-26 Hanger & Co Ltd J E Artificial hand
US20130253606A1 (en) * 2012-03-20 2013-09-26 In Chan Youn Peripheral nerve interface system and method for prosthetic hand control
CN104665828A (en) * 2013-11-27 2015-06-03 中国科学院深圳先进技术研究院 System and method based on electromyographic signal controlling remote controller
CN105615890A (en) * 2015-12-24 2016-06-01 西安交通大学 Angle and myoelectricity continuous decoding method for human body lower limb walking joint
CN105963100A (en) * 2016-04-19 2016-09-28 西安交通大学 Patient movement demand-based assistance lower limb rehabilitation robot self-adaptation control method
CN106236336A (en) * 2016-08-15 2016-12-21 中国科学院重庆绿色智能技术研究院 A kind of myoelectric limb gesture and dynamics control method
CN106923942A (en) * 2017-02-15 2017-07-07 上海术理智能科技有限公司 Upper and lower extremities motion assistant system based on the control of human body electromyographic signal

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
JIANGCHENG CHEN ET AL: ""Surface EMG based continuous estimation of human lower limb joint angles by using deep belief networks"", 《BIOMEDICAL SIGNAL PROCESSING AND CONTROL》 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109259739A (en) * 2018-11-16 2019-01-25 西安交通大学 A kind of myoelectricity estimation method of wrist joint motoring torque
CN110801226A (en) * 2019-11-01 2020-02-18 西安交通大学 Human knee joint moment testing system method based on surface electromyographic signals and application
CN110827987B (en) * 2019-11-06 2021-03-23 西安交通大学 Myoelectricity continuous prediction method and system for wrist joint torque in multi-grabbing mode
CN110827987A (en) * 2019-11-06 2020-02-21 西安交通大学 Myoelectricity continuous prediction method and system for wrist joint torque in multi-grabbing mode
CN111374808A (en) * 2020-03-05 2020-07-07 北京海益同展信息科技有限公司 Artificial limb control method and device, storage medium and electronic equipment
CN111616848A (en) * 2020-06-02 2020-09-04 中国科学技术大学先进技术研究院 Five-degree-of-freedom upper arm prosthesis control system based on FSM
CN111616848B (en) * 2020-06-02 2021-06-08 中国科学技术大学先进技术研究院 Five-degree-of-freedom upper arm prosthesis control system based on FSM
WO2022001771A1 (en) * 2020-06-29 2022-01-06 京东科技信息技术有限公司 Artificial limb control method, device and system and storage medium
CN111920416A (en) * 2020-07-13 2020-11-13 张艳 Hand rehabilitation training effect measuring method, storage medium, terminal and system
CN111920416B (en) * 2020-07-13 2024-05-03 张艳 Hand rehabilitation training effect measuring method, storage medium, terminal and system
CN112587242A (en) * 2020-12-11 2021-04-02 天津大学医疗机器人与智能系统研究院 Surgical robot master hand simulation method, master hand and application
CN114224577A (en) * 2022-02-24 2022-03-25 深圳市心流科技有限公司 Training method and device for intelligent artificial limb, electronic equipment, intelligent artificial limb and medium
CN117953413A (en) * 2024-03-27 2024-04-30 广东工业大学 Electromyographic signal validity judging method, electromyographic signal validity judging device, electromyographic signal validity judging equipment and storage medium

Also Published As

Publication number Publication date
CN109009586B (en) 2020-07-28

Similar Documents

Publication Publication Date Title
CN109009586A (en) A kind of myoelectricity continuous decoding method of the man-machine natural driving angle of artificial hand wrist joint
CN108785997B (en) Compliance control method of lower limb rehabilitation robot based on variable admittance
CN106067178B (en) A kind of continuous estimation method of hand joint movement based on muscle synergistic activation model
Chen et al. Multiple hand gesture recognition based on surface EMG signal
CN108743223B (en) Intelligent exoskeleton rehabilitation manipulator and method based on healthy side biological electric control
CN109222969A (en) A kind of wearable human upper limb muscular movement fatigue detecting and training system based on Fusion
Jiang et al. A method of recognizing finger motion using wavelet transform of surface EMG signal
CN102499797B (en) Artificial limb control method and system
CN202288542U (en) Artificial limb control device
CN112057040B (en) Upper limb movement function rehabilitation evaluation method
CN101036601A (en) Real time control device and control method by two-degrees-of freedom myoelectricity artificial hand
CN107126303A (en) A kind of upper and lower extremities exercising support method based on mobile phone A PP
Tang et al. Continuous estimation of human upper limb joint angles by using PSO-LSTM model
CN111803099A (en) Human upper limb muscle force prediction device and method based on radial basis function neural network
CN114897012A (en) Intelligent prosthetic arm control method based on vital machine interface
Bai et al. Multi-channel sEMG signal gesture recognition based on improved CNN-LSTM hybrid models
Kumar et al. A critical review on hand gesture recognition using semg: Challenges, application, process and techniques
CN104665828A (en) System and method based on electromyographic signal controlling remote controller
KR100994408B1 (en) Method and device for deducting pinch force, method and device for discriminating muscle to deduct pinch force
Wang et al. EMG signal classification for myoelectric teleoperating a dexterous robot hand
Meng et al. Estimation of ankle joint continuous motion based on electromyographic signals
Sun et al. SEMG based continuous estimation of wrist joint angle using BP neural network
Zhang et al. Research on gesture definition and electrode placement in pattern recognition of hand gesture action SEMG
Yang et al. Detection and identification the human surface Electromyogra signal
Peng et al. Evaluation of sEMG-based feature extraction and effective classification method for gait phase detection

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
GR01 Patent grant
GR01 Patent grant