CN113171214A - Multi-path feedback myoelectricity control prosthetic hand based on self-adaptive enhanced classifier and method - Google Patents
Multi-path feedback myoelectricity control prosthetic hand based on self-adaptive enhanced classifier and method Download PDFInfo
- Publication number
- CN113171214A CN113171214A CN202110584128.7A CN202110584128A CN113171214A CN 113171214 A CN113171214 A CN 113171214A CN 202110584128 A CN202110584128 A CN 202110584128A CN 113171214 A CN113171214 A CN 113171214A
- Authority
- CN
- China
- Prior art keywords
- joint
- finger
- feedback
- motor
- prosthetic hand
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 230000033001 locomotion Effects 0.000 claims abstract description 59
- 230000009471 action Effects 0.000 claims abstract description 51
- 230000003993 interaction Effects 0.000 claims abstract description 24
- 230000003183 myoelectrical effect Effects 0.000 claims abstract description 21
- 238000001228 spectrum Methods 0.000 claims abstract description 21
- 238000012216 screening Methods 0.000 claims abstract description 14
- 210000003811 finger Anatomy 0.000 claims description 47
- 210000001503 joint Anatomy 0.000 claims description 37
- 239000011159 matrix material Substances 0.000 claims description 32
- 230000005540 biological transmission Effects 0.000 claims description 19
- 210000000811 metacarpophalangeal joint Anatomy 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 14
- 210000003813 thumb Anatomy 0.000 claims description 13
- 210000001145 finger joint Anatomy 0.000 claims description 12
- 230000003044 adaptive effect Effects 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 8
- 210000004932 little finger Anatomy 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 208000031361 Hiccup Diseases 0.000 claims description 4
- 230000008030 elimination Effects 0.000 claims description 4
- 238000003379 elimination reaction Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 abstract description 7
- 238000012549 training Methods 0.000 description 17
- 230000004913 activation Effects 0.000 description 14
- 238000002567 electromyography Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 10
- 230000036544 posture Effects 0.000 description 9
- 230000000694 effects Effects 0.000 description 7
- 238000005070 sampling Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 230000003595 spectral effect Effects 0.000 description 5
- 230000009466 transformation Effects 0.000 description 5
- 238000013145 classification model Methods 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 210000003414 extremity Anatomy 0.000 description 4
- 210000003205 muscle Anatomy 0.000 description 4
- 239000000463 material Substances 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 239000011664 nicotinic acid Substances 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 238000010187 selection method Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000003064 k means clustering Methods 0.000 description 2
- 230000009023 proprioceptive sensation Effects 0.000 description 2
- 238000004804 winding Methods 0.000 description 2
- 238000002266 amputation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 210000004247 hand Anatomy 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
- 210000000707 wrist Anatomy 0.000 description 1
Images
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
- A61F2/54—Artificial arms or hands or parts thereof
- A61F2/58—Elbows; Wrists ; Other joints; Hands
- A61F2/583—Hands; Wrist joints
-
- 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
- A61F2/54—Artificial arms or hands or parts thereof
- A61F2/58—Elbows; Wrists ; Other joints; Hands
- A61F2/583—Hands; Wrist joints
- A61F2/586—Fingers
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- A61F2002/6827—Feedback system for providing user sensation, e.g. by force, contact or position
Abstract
The invention discloses a multi-path feedback myoelectricity control prosthetic hand based on a self-adaptive enhancement classifier and a method thereof, wherein a prosthetic hand body comprises a plurality of simulated fingers which are independently moved and arranged on a palm platform, and the simulated fingers are controlled to move by a motor driving module; the signal acquisition module is used for acquiring myoelectric signals of a person wearing the prosthetic hand body; the classification module is used for extracting action recognition characteristics from the electromyographic signals through short-time Fourier transform and logarithmic spectrum images, performing integrated screening on the action recognition characteristics, and then performing electromyographic classification by adopting a self-adaptive enhancement classifier to obtain a motion control instruction; the motor driving module is used for driving the finger-imitating action according to the motion control instruction; the feedback module is used for controlling the execution mode of the motor after the finger-imitating action according to the interaction force with the object, the working current of the motor and the joint pose during the finger-imitating action. Under the condition of ensuring high accuracy, the calculation amount is reduced to the maximum extent, and the speed and the accuracy of the artificial hand motion recognition are improved.
Description
Technical Field
The invention relates to the technical field of rehabilitation robots, in particular to a multi-channel feedback myoelectric control prosthetic hand based on a self-adaptive enhancement classifier and a method.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
Wearing a bionic prosthetic hand is the most direct and effective compensation method for hand amputees. The appearance of the bionic artificial hand is basically similar to that of a human hand; the structure simulates the physiological structure of the palm and the five fingers, and the motor replaces muscles to drive the movement of finger joints; basically, a simple grip or other action can be realized.
However, to the inventor's knowledge, existing prosthetic hand (not including robotic arms) control has far failed to achieve the level of flexible control. The single artificial hand is controlled by single degree of freedom, and the motion mode is only opened and closed, which is far from the real flexible control; however, the artificial limb with multiple degrees of freedom is designed to be a combination of an artificial hand and a mechanical arm, or the artificial hand is arranged on a huge base, so that a patient can hardly really wear the heavy artificial hand to operate, and particularly for the patient with an amputation at the wrist, the patient does not need a complex mechanical arm, but needs the artificial hand with both smart control and light weight design.
For the real-time control problem, the existing classification or regression based algorithm needs to make a selection on accuracy and timeliness. When the signal slides the window, the information contained is more comprehensive and the accuracy is higher if the window length is larger, but the time delay is larger, and vice versa. At present, no excellent algorithm can simultaneously realize extremely high accuracy and timeliness.
Disclosure of Invention
In order to solve the problems, the invention provides a multi-path feedback myoelectricity control prosthetic hand based on a self-adaptive enhancement classifier and a method thereof, which are used for collecting motion intention signals of a prosthetic hand wearer in real time, controlling the motion of the prosthetic hand by combining feedback information of interaction force with an object, motor working current and joint position and posture, realizing the operation requirement in daily life, realizing flexible control and quick response and improving the effect of user experience.
In order to achieve the purpose, the invention adopts the following technical scheme:
in a first aspect, the invention provides a multi-path feedback myoelectricity control prosthetic hand based on a self-adaptive enhancement classifier, which comprises a prosthetic hand body, a motor driving module, a signal acquisition module, a classification module and a feedback module, wherein the motor driving module is connected with the signal acquisition module;
the artificial hand body comprises a plurality of simulated fingers which are independently moved and arranged on the palm platform, and the simulated fingers are controlled to move by the motor driving module;
the signal acquisition module is used for acquiring myoelectric signals of a person wearing the prosthetic hand body;
the classification module receives the electromyographic signals, is configured to extract action recognition characteristics of the electromyographic signals through short-time Fourier transform and logarithmic spectrum images, performs integrated screening on the action recognition characteristics, and performs electromyographic classification by adopting a self-adaptive enhancement classifier to obtain a motion control instruction;
the motor driving module is configured to drive finger-imitating actions according to motion control instructions;
the feedback module is configured to control an execution mode of the motor after the finger-imitating action according to the interaction force with the object during the finger-imitating action, the working current of the motor and the joint pose.
In a second aspect, the present invention provides a control method based on the prosthetic hand, including:
acquiring myoelectric signals of a wearer of the prosthetic hand body;
obtaining a motion control instruction according to the classification of the electromyographic signals, and driving the finger-imitating action according to the motion control instruction; extracting action recognition characteristics from the electromyographic signals through short-time Fourier transform and logarithmic spectrum images, and performing electromyographic classification by adopting a self-adaptive enhanced classifier after performing integrated screening on the action recognition characteristics;
and acquiring the interaction force with the object, the working current of the motor and the joint pose during the finger-imitating action so as to control the execution mode of the motor after the finger-imitating action.
Compared with the prior art, the invention has the beneficial effects that:
the invention realizes the fully-driven artificial hand with high degree of freedom, wherein 14 degrees of freedom (2 thumbs and 3 other four fingers) are independently driven by corresponding encoder motors, thereby providing more flexible and accurate artificial hand control for a user, solving the design and control problems of the flexible multi-degree-of-freedom artificial hand with multi-path feedback, being different from the multi-degree-of-freedom artificial hand and needing no complex base for supporting.
The invention extracts features through short-time Fourier transform and logarithmic spectrum images, selects integrated features and then uses the self-adaptive enhancement classifier to carry out electromyographic classification, thereby reducing the calculated amount to the maximum extent and improving the speed and the accuracy of the artificial hand motion recognition under the condition of ensuring high accuracy.
The myoelectric arm ring is used as a main signal source for controlling the artificial hand, auxiliary control is carried out through the force sensor, the motor encoder and the current feedback path, and the interaction force, the joint pose and the working current of the artificial hand are fed back in real time in the motion process, so that the applicability and the safety of the gripping action are improved, and the using effect of the artificial hand is improved.
Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
Fig. 1 is a block diagram of a multi-path feedback control system of a prosthetic hand according to embodiment 1 of the present invention;
fig. 2 is an overall structure of a prosthetic hand provided in embodiment 1 of the present invention;
FIG. 3 is a schematic structural view of a prosthetic finger joint according to embodiment 1 of the present invention;
fig. 4 is a flowchart of electromyographic feature extraction and classification provided in embodiment 1 of the present invention;
fig. 5 is a block diagram of an integrated electromyography feature screening process provided in embodiment 1 of the present invention;
FIG. 6 is a flow chart of a feedback-controlled gripping motion provided in embodiment 1 of the present invention;
the artificial hand comprises a palm platform 1, a palm platform 2, an artificial hand thumb 3, an artificial hand index finger 4, an artificial hand middle finger 5, an artificial hand ring finger 6, an artificial hand little finger 7, a direct current motor 8, a driving gear 9, a transmission gear 10, a joint driven shaft 11, an inter-digital joint with a shaft 12, a groove inter-digital joint 13, a metacarpophalangeal joint transmission shaft 14, a fingertip joint with a shaft 15 and a groove fingertip joint.
The specific implementation mode is as follows:
the invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise, and it should be understood that the terms "comprises" and "comprising", and any variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
The embodiments and features of the embodiments of the present invention may be combined with each other without conflict.
Example 1
The embodiment provides a multi-path feedback myoelectricity control prosthetic hand based on a self-adaptive enhancement classifier, which comprises a prosthetic hand body, a motor driving module, a signal acquisition module, a classification module and a feedback module;
the artificial hand body comprises a plurality of simulated fingers which are independently moved and arranged on a palm platform, and the simulated fingers are controlled to move by a motor driving module;
the signal acquisition module is used for acquiring myoelectric signals of a person wearing the prosthetic hand body;
the classification module receives the electromyographic signals, is configured to extract action recognition characteristics of the electromyographic signals through short-time Fourier transform and logarithmic spectrum images, performs integrated screening on the action recognition characteristics, and performs electromyographic classification by adopting a self-adaptive enhancement classifier to obtain a motion control instruction;
the motor driving module is configured to drive finger-imitating actions according to motion control instructions;
the feedback module is configured to control an execution mode of the motor after the finger-imitating action according to the interaction force with the object during the finger-imitating action, the working current of the motor and the joint pose.
As shown in fig. 1, a block diagram of a prosthetic hand system is shown, in a training stage, an electromyographic signal is used as a data set for training a classifier, and in an actual grasping operation stage, the electromyographic signal is used as a control signal source of the classifier; the motor driving module outputs preset motion commands to each encoder motor according to the motion control commands, and controls the motors to move to the specified positions, namely, the motor driving module drives the finger-imitating action;
meanwhile, the artificial hand also considers the adaptive interaction with the object when gripping, which is mainly realized by the multi-path feedback on the artificial hand; because the gripping posture is preset, is universally used for all interactive objects and is limited by the muscle retention condition of the stump of a user, the effective myoelectricity activation pattern which can be generated by the gripping posture is not enough to be distributed to objects with different shapes, qualities and textures, and certain feedback control is needed to ensure that the gripping is safer and more reliable; the motion posture of the artificial hand is fed back by a winding encoder in the motor in real time, the working current of the motor is realized by a current feedback circuit of a motor driving module, and the interaction between the artificial hand and a gripping object is measured and fed back by a force sensor.
The inner side of the prosthetic hand is pasted with a covering layer made of flexible materials to increase the contact area with an object, meanwhile, a piezoelectric sensor is arranged between the covering layer and the shell of the prosthetic hand, when the contact force of the prosthetic hand is too large due to interference between the objects in the motion process, in order to prevent the object from being gripped and drive the motor, the feedback force enables the prosthetic hand motor to enter a hiccup control mode, the driving system is prevented from being damaged by strong current generated by motor stalling under the condition of maintaining certain interaction force, and the interference caused by environmental contact in the range of a non-force sensor or the self-reason of the prosthetic hand can be stopped by directly sending an instruction to an upper computer through a current feedback module in a motor driving module; for the grasping condition without interaction, for example, the prosthetic hand makes various gestures, and at the moment, the prosthetic hand does not intervene in force feedback, and the prosthetic hand is considered to complete the movement task after moving to the specified pose.
As shown in fig. 2, the artificial hand comprises a palm platform 1, an artificial hand thumb 2, an artificial hand index finger 3, an artificial hand middle finger 4, an artificial hand ring finger 5 and an artificial hand little finger 6; 5 direct current motors fixed in the metacarpophalangeal joints are fixed on the inner side of the palm through a fixing structure in the palm platform 1, and the direct current motors of the other 9 fingertip joints are fixed in similar structures in the fingertip joints; the flexible material covers the inner side surface of the artificial hand to increase the contact area with an object during gripping and realize balanced stress, and the pressure sensors are in strip shapes and are respectively fixed below the covering layer from fingertips to metacarpophalangeal joints to measure the interaction of force between the hand and the object during gripping.
The corresponding positions of five fingers on a palm platform are sequentially provided with a connecting thumb metacarpophalangeal joint, an index finger metacarpophalangeal joint, an interphalangeal joint, a middle finger metacarpophalangeal joint, an interphalangeal joint, a ring finger metacarpophalangeal joint, an interphalangeal joint and an interphalangeal joint, wherein the artificial thumb corresponds to two encoder direct current motors to respectively control the thumb metacarpophalangeal joint and the interphalangeal joint, the rest four fingers correspond to three encoder direct current motors to respectively control the movements of the metacarpophalangeal joint, the near-end interphalangeal joint and the far-end interphalangeal joint, and the total twelve encoder motors carry out movement control through mutually independent driving modules.
As shown in fig. 3, the palm platform 1, the artificial hand thumb 2, the direct current motor 7, the driving gear 8, the transmission gear 9, the joint driven shaft 10, the interphalangeal joint with shaft 11, the interphalangeal joint with groove 12, the metacarpophalangeal joint transmission shaft 13, the fingertip joint with shaft 14 and the fingertip joint with groove 15; the structure of an index finger 3 of a prosthetic hand, the structure of a middle finger 4 of the prosthetic hand, the structure of an ring finger 5 of the prosthetic hand and the structure of a little finger 6 of the prosthetic hand are completely consistent, the structures of interphalangeal joints of a thumb and finger joints of the other four fingers are completely consistent, the prosthetic finger joints are formed by combining an interphalangeal joint 11 with a shaft and a groove interphalangeal joint 12, a direct current motor is arranged in each joint to drive the next-stage joint to move, and each finger joint is formed by combining an interphalangeal joint 14 with a shaft and a groove finger joint;
the palm platform is provided with a near-end connecting protruding shaft at one side of the five fingers, the near-end connecting protruding shaft of the five fingers is coaxially connected with the tail end of the corresponding position of the finger of the artificial limb hand to realize rotary motion, and a transmission gear is nested on the protruding shaft;
the artificial finger units of the index finger, the middle finger, the ring finger and the little finger of the artificial hand comprise fingertip joint protruding components, fingertip joint embedding components, middle joint protruding components, middle joint embedding components, near-end joint protruding components, near-end joint embedding components, power gears, transmission gears and driving shafts; the fingertip joint is meshed with the transmission gear on the same horizontal plane through the driving shaft locked with the fingertip joint, the transmission (bevel) gear is meshed with the power (bevel) gear at an angle of 90 degrees, and the torque generated by the motor is used for driving the fingertip joint;
a motor for driving the fingertip joint is positioned in the inner space of the middle joint, a motor for driving the middle fingertip joint is positioned in the inner space of the near-end fingertip joint, a motor for driving the near-end fingertip joint is positioned in the palm platform, and the fingertip joint is used as the tail end of a finger and is not provided with a motor;
the artificial hand thumb only has a fingertip joint and an interphalangeal joint, and comprises a fingertip joint protruding component, a fingertip joint embedding component, a near-end joint protruding component, a near-end joint embedding component, a power gear, a transmission gear and a driving shaft, and the movement and transmission mode of the artificial hand thumb is the same as that of the rest four fingers.
When the encoder rotating motor moves, the power of the encoder rotating motor is transmitted to the transmission gear embedded in the finger joint through the bevel gear on the encoder rotating motor, the transmission gear drives the driving shaft meshed with the encoder, the driving shaft is integrally printed with a structure meshed with the transmission gear, and the driving shaft is structurally locked with the next finger joint, so that the driving shaft can drive the next finger joint to rotate, bend or extend.
Taking the index finger of the artificial hand in fig. 3 as an example, the interphalangeal joints are embedded with the palm platform 1 through the joint driven shaft 10, meanwhile, the interphalangeal joints and the joint driven shaft are locked, and the design of the other joints is also the same; when the direct current motor 7 works, the driving gear 8 drives the transmission gear 9 to move, then torque is transmitted to the joint driven shaft 10, the driven shaft and the interphalangeal joint are locked, and finally the joint driven shaft 10 drives the whole interphalangeal joint to realize the rotary motion of the joint.
As shown in fig. 4, which is a flow chart of electromyography (emg) feature extraction and classification, Surface electromyography (sEMG) signals are widely used in the fields of various rehabilitation devices, human-computer interaction, clinical and biomedical fields, etc.; in the classification of control commands for rehabilitation devices, in particular prosthetic hands, sEMG signals are the mainstream choice for non-invasive devices at present, and electrical signals reflecting different hand muscle activities can be recorded under non-invasive conditions;
the embodiment provides a medical expert system based on sEMG signals for classifying the gripping action induced myoelectricity of a user so as to enhance the action classification accuracy in the use of a prosthetic hand, and the medical expert system mainly comprises the following 3 modules: spectrum-Based Logarithmic transformation image Signal (LSGS), feature extraction techniques, and an adaptive enhanced k-means (AdaBoost k-means, AB-k-means) classifier.
In the actual use process, the collected sEMG data of the user is preprocessed in a way the same as that of a training data set, and then the data points collected in real time are subjected to the same Hamming sliding window processing and short-time Fourier transform to obtain a Laplace matrix and characteristic values thereof and calculate related statistical characteristics; as the integrated feature selection is carried out on the statistical features with better effect in the training stage, the step is not repeated when in actual use, and the obtained feature values are directly input into the trained self-adaptive k-means motion classifier to obtain the motion instruction.
In the embodiment, short-time Fourier transform after sliding window is performed on EMG, a spectrum image and a weight matrix thereof are obtained after logarithmic processing is performed on a frequency density spectrum, and a Laplace feature matrix and a feature vector thereof are further obtained, so that various statistical parameter features are calculated, finally, the selected classification features need three methods of chi-square feature selection, mutual information feature selection and recursive feature elimination, and the most effective data features are obtained through a Fisher formula; and finally, classifying by using a self-adaptive enhanced k-means classifier, classifying and identifying the features extracted from the surface myoelectric signals, and combining the optimized k-means classification results of a plurality of linear weak classifiers into a strong classifier by the algorithm.
Specifically, simple preprocessing is carried out on the sEMG signal, and the sEMG signal is filtered and subjected to noise reduction to be used as an original signal for motion mode classification; because the physiological signal is usually unstable and nonlinear, an LSGS (local Strand Meter-spline) model is applied to analyze the processed electromyographic signal to obtain an activation pattern with fixed characteristics, and the characteristics are extracted; collecting the extracted features with higher weight into a feature extraction algorithm; finally, in the classification stage, a novel classification model AB-k-means is adopted to map the selected sEMG characteristics to different gripping poses, so that a final classification result is realized; the classifier adopted in the embodiment has fewer learned features, so that the operation amount and the hardware requirement are reduced, and more accurate results can be provided in shorter training time.
For the non-stationary signals with high dimension of EMG, the embodiment uses the spectrogram in logarithmic domain to perform time-frequency conversion on the electromyographic signals for further processing; the specific operation of Time-frequency conversion is that Short-Time Fourier Transform (STFT) is adopted to perform Short-Time Fourier Transform on data, windowing processing needs to be performed on the data, and a Hamming window containing 256 continuous sampling points is selected as a unit of Fourier Transform; each time, the window is shifted by half of the window length, namely 128 sampling points, to completely cover the electromyogram so as to generate a spectrogram, wherein x represents the input of an original electromyogram signal, K represents a label of discrete time, z represents a Hamming window function selected by calculation, and K is a constant representing the window length and is 256;
therefore, X | f | is obtained in each frequency range of each window signal, the spectrum value is a complex number, and the logarithmic spectrum density S (f, j) is obtained after the module value of the complex number is calculated; in the calculation process of the log spectral density, marking the log spectral density by a time frame j is needed to realize the calculation;
the next step is the plotting of the logarithmic spectral density map, as shown in equation (1):
IS(f,j)＝log(S(f,j))(1)
and (3) carrying out nonlinear scaling on the log spectral density by using a log function, simultaneously enhancing the spectral components of the frequency band of the EMG signal, and obtaining a two-dimensional log spectrogram IS (f, j) in frequency and time after transforming by using the log function.
For a single task training, thousands of data sampling points may be recorded by several seconds of actions, and after short-time fourier transform, logarithmic spectrum density calculation and spectrogram drawing with time labels are performed on the data sampling points, an IS (f, j) with high resolution and containing motion action identification feature information IS obtained; in order to facilitate model training and later-stage motion recognition, the spectrogram is reduced to a resolution of 50 × 50 by a uniform sampling method, and the image is sensitive to frequencies of different actions and STFT amplitudes corresponding to the frequencies, and can be used as a reliable criterion for action classification.
The log spectrogram after dimensionality reduction iS denoted as iS (f, j), which iS a constant matrix with a resolution of 50 × 50, and since dimensionality reduction iS performed on the time variable and the frequency variable, the variables f and j no longer represent the scale of frequency and time, and therefore are only represented as algebraic expressions on the abscissa and the ordinate.
Further, the log spectrum image iS (f, j) iS changed from a 50 × 50 matrix to a 1 × 2500-scale row matrix G in a row-by-row arrangement, as shown in formula (2):
G＝[g1,g2,g3,......gn],g50(f-1)+j＝iS(f,j) (2)
the spectrum image loses the two-dimensional resolution, and adjacent elements in G are adjacent elements in the original two-dimensional log spectrum image or an end element of one line and a first element of the next line; it is clear that in the generation of the weighting matrix, there should be a difference in the weight calculation between the two cases, so that a corresponding equal-scale index matrix v, v of G is generated for each element vnAll represent corresponding to GnTwo-dimensional coordinates of (a).
And then, applying a Gaussian kernel weighting function to calculate the weights among different nodes, wherein the weights are expressed by the following formulas (3) to (4):
where the exp function is an exponential function with the natural logarithm e as the base number, dist (v)i,vj) To solve v in density spectrogrami、vjOf between, dist (g)i,gj) Are the physical distances of the elements in the image signal G, and the parameters ψ, h, ω are empirically set to 5.01, 0.2, 0.3, respectively.
For the case where the Euclidean or physical distance of an element is greater than a threshold h, when i ═ j, pi,j＝qi,j1 is ═ 1; otherwise pi,j＝qi,j0; the mutual influence is defined through the position relation of pixel points of the two frequency spectrogram, and then the final weighting matrix W is obtained through the multiplication of a formula (5):
wi,j＝pi,j×qi,j (5)
under the condition that a weighting matrix W is known, a diagonal matrix D derived from the weighting matrix W is obtained through elementary transformation, and a Laplace matrix l is obtained and is a difference matrix of subtracting W from D.
The transformation of the spectrum image is realized by performing eigenvalue decomposition on the Laplace matrix l, which needs to solve all eigenvalues of l by means of mathematical software, and sort the eigenvalues according to the size to obtain an N-dimensional eigenvalue matrix X ═ χ1,χ2,χ3,......χnThe eigenvalue diagonal matrix Λ of N, the Laplace matrix of the graph is a semi-positive definite matrix, so χ10, and for arbitrary χnAnd the characteristic value decomposition result L is expressed by the formula (6):
L＝XΛXT (6)
wherein, XTRepresenting a matrix of eigenvalues columns obtained by a matrix transposition of the eigenvalue row matrix.
As shown in fig. 5, the motion recognition features obtained in the above steps are selected, and various advantages and disadvantages of the feature value selection methods are embodied in many factors such as data type selection, accuracy, time, cost, and the like. Therefore, the present embodiment adopts an integrated feature selection method, and simultaneously adopts chi-square feature selection, mutual information, and recursive feature elimination to perform overall feature screening to obtain a feature set.
To reduce the dimensionality of the eigenvalue vector extracted from G, a set of common statistical features is chosen, namely: a first quartile, a second quartile, a standard deviation, a mean, a minimum, a mode, a maximum, a median, a range, a coefficient of variation, a skewness, a kurtosis, a trend, a sequence correlation, a self-similarity, a periodicity, a root mean square, a percentile; all the parameters are tested by respectively adopting a chi-square feature selection method, a mutual information method and a recursive feature elimination method, and the final features are subjected to threshold screening based on a Fisher criterion to obtain the most effective classification features, as shown in a formula (7):
wherein p isk，σk，ukRespectively representing the mean value, the variance and the ratio of the characteristic in the kth class in the final characteristic to be screened; the Fisher judgment rate is obtained and used for setting a threshold value for screening the final characteristics, and the formula (8) is shown:
wherein, the value of alpha is set to 0.75, rho is the number of features to be finally reserved, and v represents the feature corresponding to the threshold value; it should be noted that too few features may result in insufficient information for classification, while too many features may increase the amount of calculation and may also have an effect of overfitting the final result.
In order to classify and identify the extracted and screened features, the method further improves the accuracy of classification by adopting a self-adaptive enhancement (AdaBoost) iterative algorithm and a k-means clustering classification method; specifically, AdaBoost is responsible for training weak classifiers to identify EMG-related hand-grasp gestures, while k-means is used to learn classification models to differentiate between different hand-grasp gestures; when the k-means classification has misjudgment, AdaBoost modifies the weight of the k-means classification according to the error rate, so as to obtain an optimal single weak classifier.
The clustering principle of k-means is as follows: firstly, selecting K points in a classification space as an initial centroid; assigning each object to the centroid closest to it; obtaining a new centroid of the area according to the distribution result; repeating the above steps until the centroid is not changed any more.
The self-adaptive reinforced classifier is a machine learning algorithm used for training a group of weak classifiers, the classifier mechanism of the self-adaptive reinforced classifier is to update the weight and error rate of a training sample after each iteration, the training weight of a training set for the wrong classification is increased, the weight of a training set for the correct classification is reduced, and the effective combination of the weak classifiers is realized through the iteration of negative feedback so as to design a strong classifier, such as a formula (9):
wherein h isiRepresents the ith weak classifier, thetaiIs a pre-set performance threshold to prevent the program from falling into an infinite loop, v is a feature vector, fiRepresents the f-order component of the feature vector v and s represents the regularization parameter.
First, two sets of labeled training set data M are combined1、M2Input into the trainer, wherein each set of data comprises (x)1,y1),(x2,y2).....(xn,yn) (ii) a x represents the EMG characteristic set which is kept input after characteristic screening; y represents its corresponding tag; presetting s as 0.1, initializing weight matrix wijInitializing a cycle count parameter i to 1, and j to 1;
the weight matrix is then normalized as shown in equation (10):
a simple weak classifier is defined, as in equation (11):
h(v,fi,θi,s)＝aδ(vf>θi)+b (11)
wherein, δ represents a step function, when the f-order component of the feature vector v is greater than a preset performance threshold, the function value is constantly 1, otherwise, the function value is constantly zero; a and b are adjustment coefficients used to correct errors;
the classifier is a simple linear function, has weak classification effect, and applies a weighting matrix wijTraining k-means clustering, and obtaining the error related parameters a, b and f of the clustering resulti、θiIs expressed as in equation (12):
wherein the content of the first and second substances,
is the clustering result after regularization, and the value range is [1, -1](ii) a At error EiOn the premise of minimization, the parameters a, b and f which best meet the current weak classifier are foundi、θiTo achieve the best classification effect; and traversing each stage of the feature vectors from i-1 to i-n to obtain the best-performance weak classifier of each stage of the feature vectors, and obtaining the final strong classifier by using a formula (9).
The surface electromyography electrodes used in this embodiment perform EMG signal acquisition at a sampling frequency of 500Hz, then select a butterworth band-pass filter to filter noise signals lower than 15Hz and higher than 500Hz, and simultaneously perform 50Hz power frequency notch on the electromyography signals in order to eliminate interference of a power supply line.
The user needs to collect myoelectric control data before using the prosthetic hand, is familiar with the control relation between different myoelectric activation modes and the prosthetic hand, trains the accuracy of a classification model, and dynamically selects the number of effective actions according to the muscle persistence condition of the residual limb of the user, and each grasping action needs to be matched with a specific myoelectric activation mode, so that each myoelectric activation mode is required to be independent and repeatable and can be actively controlled by the user; the repetitive training of each myoelectricity activation mode allows a user to determine the speed and the strength of myoelectricity activation by himself as long as the repeatability of movement is ensured.
In the exercise process, a user inputs basic information and trains the basic information, then myoelectricity activation of the stump of the user is controlled to be used as a main signal source for exercise control, and the screened features are classified by a self-adaptive enhanced k-means classifier and then output as a final control instruction to drive the artificial hand.
After the artificial hand acts, the feedback signal is used as a gate control signal of the movement to control the movement of the artificial hand, as shown in the table 1; wherein, KE、KC、KFRepresenting encoder feedback, current feedback and interaction force feedback in the feedback signal respectivelyThe event that the machine does not move to the designated position, the current does not reach the threshold value and the force feedback does not reach the threshold value is defined as logic 0; the events that the motor reaches a specified position, the current exceeds a threshold value and the force feedback exceeds the threshold value are defined as logic 1;
TABLE 1 feedback control
There are three forms of multipath feedback: force feedback, encoder feedback, current feedback. In the artificial hand control system, the force feedback is used for preventing damage caused by overload of interaction force in the gripping process; the encoder feeds back the change of the winding resistance of the encoder to judge the number of turns of the motor, so that the rotation angle of each joint and the relative position between the joints are obtained; the current feedback is used for ensuring the most basic electrical safety of the artificial hand and preventing the artificial hand and a circuit thereof from being burnt out due to faults; the multi-path feedback can provide perception feedback similar to sense of touch and proprioception for the prosthetic hand, and can also monitor the working state of the controlled motor in real time, so that the prosthetic hand is endowed with higher reliability and flexibility;
the piezoelectric force sensors are arranged in a matrix form and are uniformly distributed in the soft prosthetic hand, each node in the matrix network is a pressure sensitive resistor, the size and the position of the resistance value change of the resistors are determined by scanning the current change of rows and columns, so that the position and the stress condition of the prosthetic hand contacting an object are determined, the piezoelectric force sensors cover the palm and the inner sides of fingers and serve as the 'skin' of the prosthetic hand, and a feedback signal similar to 'touch' is provided.
The encoder direct current motor can determine the forward rotation or reverse rotation stroke of the motor through the change of the resistance value of the encoder, and convert the stroke of the motor into the angle change of the joint rotation; the current motion posture of the prosthetic hand is determined by integrating the rotation postures of the 14 joints and is transmitted to the microcontroller and the upper computer, and the motor encoder can enable the control system to acquire the motion posture of the joint of the prosthetic hand in real time, so that the prosthetic hand is controlled to accurately move to a fixed posture, and a feedback signal equivalent to 'proprioception' is provided.
The current feedback function is integrated in the motor driving module, and the power supply is stable under the condition that the prosthetic hand normally moves, so that the current feedback reflects the running power of the motor; when the motion of the motor of the prosthetic hand is blocked, the current is increased sharply, and the control system receives overhigh feedback current to stop the motion of the prosthetic hand; the motor driving module receives signal input from the microcontroller, the command is sent to the motor driving module in a Pulse Width Modulation (PWM) signal form, the power input of the motor is provided by external power supply, and the PWM signal is converted into external power supply level in the motor driving module to drive the motor to move. And meanwhile, a current feedback interface on the module can feed the current passing through the direct current motor back to the upper computer.
The control of the gripping process by the feedback signal is shown in fig. 6, and the judgment condition in the program block diagram is to judge the feedback signal, wherein different event branches cover a typical paradigm of gripping, specifically:
(1) presetting a current threshold, a force threshold and a pose state threshold; when the artificial hand moves, the interaction force and the feedback current of the encoder are both within the threshold value, and when each joint of the artificial hand reaches the predicted movement position, the end of a normal gripping movement can be judged through the joint movement state feedback of the encoder motor.
(2) The current feedback of the prosthetic hand is higher than the threshold value in the motion process, which represents that the motor is locked up to a certain degree; if the force feedback is also above the threshold value, it can be assumed that the interaction between the prosthetic hand and the object prevents further flexion and extension of the joint, and the goal of the grip has been achieved in that, in order to protect the prosthetic hand and the gripped object, the motor enters a hiccup-controlled mode, maintaining the necessary interaction force while preventing overcurrent.
(3) The current feedback of the prosthetic hand is higher than the threshold value in the motion process, but the feedback force is low, under the condition that no interaction force exists, the current overload represents the motion fault of the prosthetic hand, at the moment, the control system stops the current motion, and the encoder motor is switched off to ensure the electrical safety.
Example 2
The embodiment provides a control method based on the prosthetic hand, which comprises the following steps:
acquiring myoelectric signals of a wearer of the prosthetic hand body;
obtaining a motion control instruction according to the classification of the electromyographic signals, and driving the finger-imitating action according to the motion control instruction; extracting action recognition characteristics from the electromyographic signals through short-time Fourier transform and logarithmic spectrum images, and performing electromyographic classification by adopting a self-adaptive enhanced classifier after performing integrated screening on the action recognition characteristics;
and acquiring the interaction force with the object, the working current of the motor and the joint pose during the finger-imitating action so as to control the execution mode of the motor after the finger-imitating action.
In the control method, basic information of a user is determined, a control system is initialized, the bionic artificial hand and the myoelectricity acquisition arm ring are worn by the user, training of a myoelectricity activation mode is carried out, a preset artificial hand action posture is set, and the selected myoelectricity activation mode needs to have specificity, spontaneity and repeatability;
mapping the effective myoelectric activation modes which can be realized by the user into the artificial limb manual work which is supposed in advance one by one, and enabling the user to be familiar with the control relation; the number of actions matched with the artificial hand wearer is adjusted according to the stump activation condition of the artificial hand wearer, and enough freedom of movement can be provided to match all possible myoelectric activation patterns.
Further, after the user is familiar with the rules for operating the prosthetic hand, a gripping test of a preset centralized action is carried out, and the classification model is trained through the obtained labeled data; selecting a logarithmic transformation image signal based on a frequency spectrum as a material for myoelectricity classification, extracting a Laplace matrix from the myoelectricity signal through sliding window short-time Fourier transformation, then obtaining each characteristic quantity of the characteristic matrix, and performing integrated characteristic selection on the characteristic quantity as input of classification; finally, selecting a self-adaptive enhanced k-means algorithm to train a plurality of linear weak classifiers to form a self-adaptive enhanced classifier which is used as a myoelectricity classification tool;
in the exercise process, a user controls myoelectric activation of the stump of the user to serve as a main signal source of exercise control, the myoelectric signals are classified and then serve as final control instructions to drive the artificial hand, the position signals of the artificial hand are monitored in real time through feedback control gripping, the exercise is finished when the artificial hand moves to a specified position, and intermittent hiccup control or motor disconnection is selected to prevent the damage of the artificial hand and a gripping object caused by exercise interference when the feedback force and the current exceed threshold values.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.
Claims (10)
1. A multi-path feedback myoelectric control prosthetic hand based on a self-adaptive enhancement classifier is characterized by comprising a prosthetic hand body, a motor driving module, a signal acquisition module, a classification module and a feedback module;
the artificial hand body comprises a plurality of simulated fingers which are independently moved and arranged on the palm platform, and the simulated fingers are controlled to move by the motor driving module;
the signal acquisition module is used for acquiring myoelectric signals of a person wearing the prosthetic hand body;
the classification module receives the electromyographic signals, is configured to extract action recognition characteristics of the electromyographic signals through short-time Fourier transform and logarithmic spectrum images, performs integrated screening on the action recognition characteristics, and performs electromyographic classification by adopting a self-adaptive enhancement classifier to obtain a motion control instruction;
the motor driving module is configured to drive finger-imitating actions according to motion control instructions;
the feedback module is configured to control an execution mode of the motor after the finger-imitating action according to the interaction force with the object during the finger-imitating action, the working current of the motor and the joint pose.
2. The multi-feedback electromyographic control prosthetic hand based on the adaptive enhanced classifier of claim 1, wherein in the classification module, the action recognition feature extraction process comprises performing short-time fourier transform after sliding window on the electromyographic signals, performing logarithm processing on the frequency density spectrum to obtain a spectrum image and a weight matrix thereof, and obtaining a laplacian feature matrix and a feature vector according to the spectrum image and the weight matrix thereof to obtain a plurality of action recognition features.
3. The adaptive classifier-based multi-feedback myoelectrically-controlled prosthetic hand according to claim 1, wherein in the classification module, the integrated screening process comprises chi-squared feature selection, mutual information feature selection and recursive feature elimination on the motion recognition features, followed by screening by fisher's formula.
4. The adaptive reinforced classifier based multi-feedback electromyographic control prosthetic hand of claim 1, wherein in the feedback module, the feedback module is configured to determine that a normal gripping motion is over if the interaction force and the motor operating current are both within a threshold range and the joints of the simulated fingers reach the expected motion positions.
5. The adaptive reinforced classifier based multi-feedback electromyographic control prosthetic hand of claim 1, wherein in the feedback module, the feedback module is configured to control the motor to enter the hiccup control mode if the interaction force and the motor working current exceed a threshold range and each joint of the simulated finger reaches a predicted motion position.
6. The adaptive reinforced classifier based multi-feedback electromyographic control prosthetic hand of claim 1, wherein in the feedback module, the feedback module is configured to control the motor to stop if the motor operating current exceeds a threshold range and the interaction force is below a threshold.
7. The multi-channel feedback myoelectric control prosthetic hand based on the self-adaptive enhanced classifier as claimed in claim 1, wherein the corresponding positions of the five fingers on the palm platform are sequentially provided with a connecting thumb metacarpophalangeal joint and interphalangeal joint, an index finger metacarpophalangeal joint and interphalangeal joint, a middle finger metacarpophalangeal joint and interphalangeal joint, a ring finger metacarpophalangeal joint and interphalangeal joint and a little finger metacarpophalangeal joint and interphalangeal joint; the artificial hand thumb corresponds to two encoder direct current motors and respectively controls the metacarpophalangeal joint and one interphalangeal joint of the thumb, and the other four fingers correspond to three encoder direct current motors and respectively control the motion of the metacarpophalangeal joint, the proximal interphalangeal joint and the distal interphalangeal joint.
8. The adaptive classifier based multi-feedback electromyographic control prosthetic hand of claim 1, wherein the prosthetic units for the prosthetic hand index finger, prosthetic hand middle finger, prosthetic hand ring finger, and prosthetic hand little finger on the palm platform each comprise a fingertip joint protruding member, a fingertip joint embedding member, a middle joint protruding member, a middle joint embedding member, a proximal joint protruding member, a proximal joint embedding member, a power gear, a transmission gear, and a drive shaft; when the motor moves, the power of the motor is transmitted to the transmission gear embedded in the finger joint through the bevel gear, the transmission gear drives the driving shaft meshed with the transmission gear, and the driving shaft is locked with the next finger joint so that the driving shaft drives the next finger joint to rotate, bend or extend.
9. The adaptive reinforced classifier based multi-feedback myoelectric control prosthetic hand according to claim 1, wherein the palm platform is provided with a proximal connecting protruding shaft at one side of the five fingers, the proximal connecting protruding shaft of the five fingers is coaxially connected with the tail end of the corresponding position of the artificial hand finger to realize the rotation movement, and a transmission gear is nested on the protruding shaft.
10. A method of controlling a prosthetic hand according to any of claims 1-9, comprising:
acquiring myoelectric signals of a wearer of the prosthetic hand body;
obtaining a motion control instruction according to the classification of the electromyographic signals, and driving the finger-imitating action according to the motion control instruction; extracting action recognition characteristics from the electromyographic signals through short-time Fourier transform and logarithmic spectrum images, and performing electromyographic classification by adopting a self-adaptive enhanced classifier after performing integrated screening on the action recognition characteristics;
and acquiring the interaction force with the object, the working current of the motor and the joint pose during the finger-imitating action so as to control the execution mode of the motor after the finger-imitating action.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110584128.7A CN113171214B (en) | 2021-05-27 | 2021-05-27 | Multi-path feedback myoelectric control prosthetic hand based on self-adaptive enhancement classifier and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110584128.7A CN113171214B (en) | 2021-05-27 | 2021-05-27 | Multi-path feedback myoelectric control prosthetic hand based on self-adaptive enhancement classifier and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113171214A true CN113171214A (en) | 2021-07-27 |
| CN113171214B CN113171214B (en) | 2023-10-24 |
Family
ID=76927267
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110584128.7A Active CN113171214B (en) | 2021-05-27 | 2021-05-27 | Multi-path feedback myoelectric control prosthetic hand based on self-adaptive enhancement classifier and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113171214B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114224578A (en) * | 2021-12-06 | 2022-03-25 | 深圳市心流科技有限公司 | Semi-physical simulation test system and method for artificial limb test |
| CN116028841A (en) * | 2022-09-01 | 2023-04-28 | 贵州大学 | Work system and method for identifying article category by tactile perception |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08294503A (en) * | 1995-04-25 | 1996-11-12 | Kenzo Akazawa | Motor-driven artificial hand |
| US20120190940A1 (en) * | 2009-06-30 | 2012-07-26 | Orthosensor | Sensored prosthetic component and method |
| US20160158033A1 (en) * | 2013-07-12 | 2016-06-09 | University Of Oregon | Powered prosthetic devices using EMG-based locomotion state classifier |
| CN106890038A (en) * | 2017-03-21 | 2017-06-27 | 上海师范大学 | Prosthetic hand control system and its control method based on MYO armlets |
| CN107049570A (en) * | 2017-03-13 | 2017-08-18 | 日照若比邻机器人科技有限公司 | Manipulator control system |
| CN107657964A (en) * | 2017-08-15 | 2018-02-02 | 西北大学 | Depression aided detection method and grader based on acoustic feature and sparse mathematics |
| CN108272537A (en) * | 2018-03-25 | 2018-07-13 | 唐山云时代网络科技有限公司 | A kind of modular multiple degrees of freedom under-actuated bionic prosthetic hand |
| CN108814778A (en) * | 2018-07-19 | 2018-11-16 | 郭伟超 | A kind of myoelectricity humanoid dexterous prosthetic hand cascade Mach-Zehnder interferometer method and system |
| CN110163049A (en) * | 2018-07-18 | 2019-08-23 | 腾讯科技（深圳）有限公司 | A kind of face character prediction technique, device and storage medium |
| CN110731839A (en) * | 2018-07-18 | 2020-01-31 | 北京理工大学 | kinds of artificial hands |
| CN110946683A (en) * | 2019-11-25 | 2020-04-03 | 西安交通大学 | Electric stimulation driving method and system for artificial finger tip pressure and joint angle feedback |
| CN111209857A (en) * | 2020-01-06 | 2020-05-29 | 武汉科技大学 | Two-way recursive myoelectric feature selection method based on myoelectric signals |
-
2021
- 2021-05-27 CN CN202110584128.7A patent/CN113171214B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH08294503A (en) * | 1995-04-25 | 1996-11-12 | Kenzo Akazawa | Motor-driven artificial hand |
| US20120190940A1 (en) * | 2009-06-30 | 2012-07-26 | Orthosensor | Sensored prosthetic component and method |
| US20160158033A1 (en) * | 2013-07-12 | 2016-06-09 | University Of Oregon | Powered prosthetic devices using EMG-based locomotion state classifier |
| CN107049570A (en) * | 2017-03-13 | 2017-08-18 | 日照若比邻机器人科技有限公司 | Manipulator control system |
| CN106890038A (en) * | 2017-03-21 | 2017-06-27 | 上海师范大学 | Prosthetic hand control system and its control method based on MYO armlets |
| CN107657964A (en) * | 2017-08-15 | 2018-02-02 | 西北大学 | Depression aided detection method and grader based on acoustic feature and sparse mathematics |
| CN108272537A (en) * | 2018-03-25 | 2018-07-13 | 唐山云时代网络科技有限公司 | A kind of modular multiple degrees of freedom under-actuated bionic prosthetic hand |
| CN110163049A (en) * | 2018-07-18 | 2019-08-23 | 腾讯科技（深圳）有限公司 | A kind of face character prediction technique, device and storage medium |
| CN110731839A (en) * | 2018-07-18 | 2020-01-31 | 北京理工大学 | kinds of artificial hands |
| CN108814778A (en) * | 2018-07-19 | 2018-11-16 | 郭伟超 | A kind of myoelectricity humanoid dexterous prosthetic hand cascade Mach-Zehnder interferometer method and system |
| CN110946683A (en) * | 2019-11-25 | 2020-04-03 | 西安交通大学 | Electric stimulation driving method and system for artificial finger tip pressure and joint angle feedback |
| CN111209857A (en) * | 2020-01-06 | 2020-05-29 | 武汉科技大学 | Two-way recursive myoelectric feature selection method based on myoelectric signals |
Non-Patent Citations (4)
| Title |
|---|
| 杨新亮等: "基于表面肌电信号的时频组合特征融合识别", 《华中科技大学学报(自然科学版)》 * |
| 杨新亮等: "基于表面肌电信号的时频组合特征融合识别", 《华中科技大学学报(自然科学版)》, vol. 39, 15 November 2011 (2011-11-15) * |
| 赵漫丹等: "基于肌电信号层级分类的手部动作识别方法", 《北京生物医学工程》 * |
| 赵漫丹等: "基于肌电信号层级分类的手部动作识别方法", 《北京生物医学工程》, vol. 33, no. 05, 15 October 2014 (2014-10-15) * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114224578A (en) * | 2021-12-06 | 2022-03-25 | 深圳市心流科技有限公司 | Semi-physical simulation test system and method for artificial limb test |
| CN116028841A (en) * | 2022-09-01 | 2023-04-28 | 贵州大学 | Work system and method for identifying article category by tactile perception |
| CN116028841B (en) * | 2022-09-01 | 2023-08-15 | 贵州大学 | Work system and method for identifying article category by tactile perception |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113171214B (en) | 2023-10-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10959863B2 (en) | Multi-dimensional surface electromyogram signal prosthetic hand control method based on principal component analysis | |
| Meattini et al. | An sEMG-based human–robot interface for robotic hands using machine learning and synergies | |
| Wang et al. | Design and myoelectric control of an anthropomorphic prosthetic hand | |
| Gijsberts et al. | Stable myoelectric control of a hand prosthesis using non-linear incremental learning | |
| Ibrahimy et al. | Design and optimization of Levenberg-Marquardt based neural network classifier for EMG signals to identify hand motions | |
| Gijsberts et al. | Movement error rate for evaluation of machine learning methods for sEMG-based hand movement classification | |
| CN110675933B (en) | Finger mirror image rehabilitation training system | |
| Strazzulla et al. | Online bimanual manipulation using surface electromyography and incremental learning | |
| Yamanoi et al. | EMG-based posture classification using a convolutional neural network for a myoelectric hand | |
| RU2635632C1 (en) | Method and system of intellectual bionic limb control | |
| Al-Faiz et al. | A k-nearest neighbor based algorithm for human arm movements recognition using EMG signals | |
| CN113171214B (en) | Multi-path feedback myoelectric control prosthetic hand based on self-adaptive enhancement classifier and method | |
| Portnova-Fahreeva et al. | Linear and non-linear dimensionality-reduction techniques on full hand kinematics | |
| CN112022619A (en) | Multi-mode information fusion sensing system of upper limb rehabilitation robot | |
| Fani et al. | Assessment of myoelectric controller performance and kinematic behavior of a novel soft synergy-inspired robotic hand for prosthetic applications | |
| Stival et al. | Online subject-independent modeling of semg signals for the motion of a single robot joint | |
| Bimbraw et al. | Towards sonomyography-based real-time control of powered prosthesis grasp synergies | |
| Ryait et al. | Study of issues in the development of surface EMG controlled human hand | |
| Fan et al. | Robust neural decoding for dexterous control of robotic hand kinematics | |
| KR100994408B1 (en) | Method and device for deducting pinch force, method and device for discriminating muscle to deduct pinch force | |
| Mohebbian et al. | A comprehensive review of myoelectric prosthesis control | |
| Ahmed et al. | Design and multichannel electromyography system-based neural network control of a low-cost myoelectric prosthesis hand | |
| JP3816762B2 (en) | Neural network, neural network system, and neural network processing program | |
| Pambudi et al. | The bionic hand movement using myo sensor and neural networks | |
| Fu et al. | Identification of finger movements from forearm surface EMG using an augmented probabilistic neural network |
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 |