阅读说明：本技术 基于fsm的五自由度上臂假肢控制系统 (Five-degree-of-freedom upper arm prosthesis control system based on FSM ) 是由 李智军 任晓前 李国欣 高洪波 黄金 皮明 于 2020-06-02 设计创作，主要内容包括：本发明提供了一种基于FSM的五自由度上臂假肢控制系统,该系统基于具有五自由度的上臂假肢,采用多通道的表面肌电信号,能够精准地识别使用者的意图；使用BP神经网络作为分类器对经过处理的肌电信号进行分类,使用基于有限状态机的编码模块对肌电信号进行二次分类,并根据进一步分类后的肌电信号产生运动状态编码,通过运动控制模块对假肢的运动做出规划并执行。本发明很好地解决传统假肢控制系统操作不方便,训练过程漫长、控制自由度少、意图识别偏差率高等一系列问题。(The invention provides a five-degree-of-freedom upper arm prosthesis control system based on an FSM (finite state machine), which is based on an upper arm prosthesis with five degrees of freedom, adopts multi-channel surface electromyographic signals and can accurately identify the intention of a user; the BP neural network is used as a classifier to classify the processed electromyographic signals, a coding module based on a finite-state machine is used to carry out secondary classification on the electromyographic signals, a motion state code is generated according to the electromyographic signals after further classification, and the motion of the artificial limb is planned and executed through a motion control module. The invention well solves a series of problems of inconvenient operation, long training process, low control freedom, high intention recognition deviation rate and the like of the traditional artificial limb control system.)

1. A five-freedom upper arm prosthesis control system based on FSM is characterized by comprising:

the system comprises a multi-channel surface electromyographic signal acquisition system and an artificial limb electromyographic control system;

the multichannel surface electromyogram signal acquisition system comprises: the signal acquisition module and the signal processing module are used for acquiring and processing the electromyographic signals of the user;

the artificial limb myoelectricity control module comprises: the device comprises an A/D conversion module, a BP neural network, a coding module, a motion control module, a man-machine interaction module, a training module and a storage module.

2. The FSM-based five degree of freedom upper arm prosthesis control system of claim 1, wherein the signal acquisition module acquires multi-channel myoelectric signals from a user's healthy limb, identifying the user's motor intention;

the signal processing module is used for amplifying, filtering and zeroing the received electromyographic signals to obtain recognizable electromyographic signals and sending the recognizable electromyographic signals to the A/D conversion module;

and the A/D conversion module is used for converting the analog signals into digital signals and sending the digital signals to the BP neural network and the training module.

3. The FSM-based five-degree-of-freedom upper arm prosthesis control system of claim 1, wherein initialization of a BP neural network is completed by reading neural network parameters, and in a use mode, myoelectric signals of a user are classified by using the initialized BP neural network, and classification results are sent to the encoding module;

the BP neural network classifies the received multi-channel electromyographic signals into a flexor longus signal L, a flexor digitalis signal S, a flexor digitalis signal P, a flexor digitalis signal D, a flexor radialis signal R and a synchronous signal C according to a BP algorithm.

4. The FSM-based five-degree-of-freedom upper arm prosthesis control system according to claim 1, characterized in that the coding module is used to calculate the time series of electromyographic signals, further classifying them according to duration, implemented by a state machine, having two working states: the wrist-hand system control state and the elbow-arm system control state are switched by the further classified electromyographic signals;

the coding module further classifies the flexor longus signal L, the flexor digitorum longus signal S, the flexor digitorum profundus signal P, the flexor digitorum signal D, the flexor radiocarpi digitorum signal R and the synchronous signal C with the duration exceeding the threshold T into a coding flexor longus signal L, a coding flexor digitorum signal S, a coding flexor digitorum profundus signal P, a coding flexor digitorum signal D, a coding flexor radiocarpi digitorum signal R and a switching synchronous signal C according to the time sequence of the signals, the short classification signal with the duration not exceeding the threshold T is coded into an invalid signal, and the artificial limb keeps the.

5. The FSM-based five-degree-of-freedom upper arm prosthesis control system of claim 1, wherein the encoding module comprises two working states, a wrist-hand system control state and an elbow-arm system control state;

when the wrist and hand system is in a wrist and hand system state, the wrist and hand system is used for controlling the wrist joint and the hand according to the further classified electromyographic signals;

when the elbow joint and the upper arm are in an elbow-arm control state, controlling the amputation of the elbow joint and the upper arm;

the electromyographic signals classified by the coding module are sent into a coding state selection stack with the depth of 2, the coding state selection stack has 36 states, 11 states are used for controlling the state switching of a wrist-hand system and an elbow-arm system, when the coding module is in the elbow-arm control state, 8 motion states exist, when the coding module is in the wrist-hand control state, 14 motion states exist, and the remaining three states enable the artificial limb to maintain the original state.

6. The FSM-based five-degree-of-freedom upper arm prosthesis control system according to claim 1, wherein the working state of the coding module is switched when the classification result of the electromyographic signals is a switching synchronization signal c, and the coding state selection stack is emptied when a state switching occurs.

7. The FSM-based five degree-of-freedom upper arm prosthesis control system of claim 1, wherein the motion control module plans the motion of each joint of the prosthesis and controls each joint to move along a planned trajectory.

8. The FSM-based five-degree-of-freedom upper arm prosthesis control system of claim 1, wherein the human-machine interaction module is configured to select a working mode, display a current working state and display a classification result.

9. The FSM-based five-degree-of-freedom upper arm prosthesis control system of claim 1, wherein the training module is configured to train myoelectric signals collected by standard movements of a user in a training mode to obtain neural network parameters.

10. The FSM-based five degree-of-freedom upper arm prosthesis control system of claim 1, wherein the storage module is configured to receive and store parameters of a BP neural network.

Technical Field

The invention relates to the field of myoelectric artificial limb control, in particular to a five-degree-of-freedom upper arm artificial limb control system based on an FSM (finite State machine).

Background

An electromyographic prosthetic limb is a rehabilitation device which uses an electromyographic signal (EMG) of a human residual limb as a control signal to control the motion of a prosthetic hand so as to realize the functions of grabbing, knocking, rotating and the like similar to human hands. The myoelectric artificial limb identifies the movement intention of a patient by collecting myoelectric signals of the patient, further controls the movement of the artificial limb, has more perfect functions than the traditional traction type artificial limb and the switch electric artificial limb, has stronger intuition and good bionic effect, and is the hot research direction of the artificial limb.

However, the conventional artificial limb control method based on state conversion usually needs to switch states through signal types and duration, and is inconvenient to operate, and in the actual use process, a series of problems of complex and long training process, less freedom, insufficient dexterity of actions, high deviation rate of intention recognition and the like exist, so that the use feeling of a user is poor, the mental stress is high, and the abandon rate of the artificial limb is high.

Disclosure of Invention

Aiming at the defects in the prior art, the invention aims to provide a five-degree-of-freedom upper arm prosthesis control system based on an FSM (finite state machine).

The invention provides a five-freedom upper arm prosthesis control system based on FSM, comprising:

the system comprises a multi-channel surface electromyographic signal acquisition system and an artificial limb electromyographic control system;

the multichannel surface electromyogram signal acquisition system comprises: the signal acquisition module and the signal processing module are used for acquiring and processing the electromyographic signals of the user;

the artificial limb myoelectricity control module comprises: the device comprises an A/D conversion module, a BP neural network, a coding module, a motion control module, a man-machine interaction module, a training module and a storage module.

Preferably, the signal acquisition module acquires multi-channel electromyogram signals from the limbs of the user to identify the movement intention of the user;

the signal processing module is used for amplifying, filtering and zeroing the received electromyographic signals to obtain recognizable electromyographic signals and sending the recognizable electromyographic signals to the A/D conversion module;

and the A/D conversion module is used for converting the analog signals into digital signals and sending the digital signals to the BP neural network and the training module.

Preferably, the initialization of the BP neural network is completed by reading the neural network parameters, the initialized BP neural network is used for classifying the electromyographic signals of the user in the use mode, and the classification result is sent to the coding module;

the BP neural network classifies the received multi-channel electromyographic signals into a flexor longus signal L, a flexor digitalis signal S, a flexor digitalis signal P, a flexor digitalis signal D, a flexor radialis signal R and a synchronous signal C according to a BP algorithm.

Preferably, the coding module is configured to calculate a time sequence of the electromyographic signals, and further classify the electromyographic signals according to duration, and the coding module is implemented by a state machine and has two working states: the wrist-hand system control state and the elbow-arm system control state are switched by the further classified electromyographic signals;

the coding module further classifies the flexor longus signal L, the flexor digitorum longus signal S, the flexor digitorum profundus signal P, the flexor digitorum signal D, the flexor radiocarpi digitorum signal R and the synchronous signal C with the duration exceeding the threshold T into a coding flexor longus signal L, a coding flexor digitorum signal S, a coding flexor digitorum profundus signal P, a coding flexor digitorum signal D, a coding flexor radiocarpi digitorum signal R and a switching synchronous signal C according to the time sequence of the signals, the short classification signal with the duration not exceeding the threshold T is coded into an invalid signal, and the artificial limb keeps the.

Preferably, the encoding module comprises two working states, a wrist and hand system control state and an elbow and arm system control state;

when the wrist and hand system is in a wrist and hand system state, the wrist and hand system is used for controlling the wrist joint and the hand according to the further classified electromyographic signals;

when the elbow joint and the upper arm are in an elbow-arm control state, controlling the amputation of the elbow joint and the upper arm;

the electromyographic signals classified by the coding module are sent into a coding state selection stack with the depth of 2, the coding state selection stack has 36 states, 11 states are used for controlling the state switching of a wrist-hand system and an elbow-arm system, when the coding module is in the elbow-arm control state, 8 motion states exist, when the coding module is in the wrist-hand control state, 14 motion states exist, and the remaining three states enable the artificial limb to maintain the original state.

Preferably, when the classification result of the electromyographic signals is the switching synchronization signal c, the working state of the coding module is switched, and when the state switching occurs, the coding state selection stack is emptied.

Preferably, the motion control module plans the motion of each joint of the prosthesis and controls each joint to move along the planned track.

Preferably, the human-computer interaction module is configured to select a working mode, display a current working state, and display a classification result.

Preferably, the training module is configured to train an electromyographic signal acquired by a standard motion performed by a user in a training mode to obtain a neural network parameter.

Preferably, the storage module is configured to receive and store the parameters of the BP neural network.

Compared with the prior art, the invention has the following beneficial effects:

the artificial limb control system realizes the state selection and the motion control of the upper arm myoelectric artificial limb with five degrees of freedom, has simple operation, high myoelectric signal classification speed, high accuracy and more controllable degrees of freedom, and can control the artificial limb to complete complex actions with better stability and control precision.

Drawings

Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:

fig. 1 is a schematic structural diagram of a five-degree-of-freedom upper arm prosthesis control system based on an FSM according to the present invention.

Fig. 2 is a schematic waveform diagram of a flexor longus signal L, a flexor digitorum superficialis signal S, a flexor digitorum profundus signal P, a extensor digitorum D, a flexor carpi radialis signal R and a synchronization signal C according to the present invention.

Fig. 3 is a schematic waveform diagram of the encoded flexor longus signal l, the encoded flexor digitorum superficialis signal s, the encoded flexor digitorum profundus signal d, the encoded flexor radiocarpi flexor r radiocarpi signal r and the switching synchronization signal c according to the present invention.

Fig. 4 is a schematic working flow diagram of a five-degree-of-freedom upper arm prosthesis control system based on an FSM according to the present invention.

Detailed Description

The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.

The invention provides a five-freedom upper arm prosthesis control system based on FSM, comprising:

the system comprises a multi-channel surface electromyographic signal acquisition system and an artificial limb electromyographic control system;

the multichannel surface electromyogram signal acquisition system comprises: the signal acquisition module and the signal processing module are used for acquiring and processing the electromyographic signals of the user;

the artificial limb myoelectricity control module comprises: the device comprises an A/D conversion module, a BP neural network, a coding module, a motion control module, a man-machine interaction module, a training module and a storage module.

Specifically, the signal acquisition module acquires multi-channel electromyographic signals from the limbs of a user and identifies the movement intention of the user;

the signal processing module is used for amplifying, filtering and zeroing the received electromyographic signals to obtain recognizable electromyographic signals and sending the recognizable electromyographic signals to the A/D conversion module;

and the A/D conversion module is used for converting the analog signals into digital signals and sending the digital signals to the BP neural network and the training module.

Specifically, the initialization of the BP neural network is completed by reading the neural network parameters, the initialized BP neural network is used for classifying the electromyographic signals of the user in the use mode, and the classification result is sent to the coding module;

the BP neural network classifies the received multi-channel electromyographic signals into a flexor longus signal L, a flexor digitalis signal S, a flexor digitalis signal P, a flexor digitalis signal D, a flexor radialis signal R and a synchronous signal C according to a BP algorithm.

Specifically, the encoding module is used for calculating a time sequence of the electromyographic signals, further classifying the electromyographic signals according to duration, is implemented by a state machine, and has two working states: the wrist-hand system control state and the elbow-arm system control state are switched by the further classified electromyographic signals;

the coding module further classifies the flexor longus signal L, the flexor digitorum longus signal S, the flexor digitorum profundus signal P, the flexor digitorum signal D, the flexor radiocarpi digitorum signal R and the synchronous signal C with the duration exceeding the threshold T into a coding flexor longus signal L, a coding flexor digitorum signal S, a coding flexor digitorum profundus signal P, a coding flexor digitorum signal D, a coding flexor radiocarpi digitorum signal R and a switching synchronous signal C according to the time sequence of the signals, the short classification signal with the duration not exceeding the threshold T is coded into an invalid signal, and the artificial limb keeps the.

Specifically, the encoding module comprises two working states, namely a wrist-hand system control state and an elbow-arm system control state;

when the wrist and hand system is in a wrist and hand system state, the wrist and hand system is used for controlling the wrist joint and the hand according to the further classified electromyographic signals;

when the elbow joint and the upper arm are in an elbow-arm control state, controlling the amputation of the elbow joint and the upper arm;

the electromyographic signals classified by the coding module are sent into a coding state selection stack with the depth of 2, the coding state selection stack has 36 states, 11 states are used for controlling the state switching of a wrist-hand system and an elbow-arm system, when the coding module is in the elbow-arm control state, 8 motion states exist, when the coding module is in the wrist-hand control state, 14 motion states exist, and the remaining three states enable the artificial limb to maintain the original state.

Specifically, when the classification result of the electromyographic signal is a switching synchronization signal c, the working state of the coding module is switched, and when the state switching occurs, the coding state selection stack is emptied.

Specifically, the motion control module plans the motion of each joint of the prosthesis and controls each joint to move along a planned track.

Specifically, the human-computer interaction module is used for selecting a working mode, displaying a current working state and displaying a classification result.

Specifically, the training module is configured to train an electromyographic signal acquired by a user performing a standard motion in a training mode to obtain a neural network parameter.

Specifically, the storage module is configured to receive and store parameters of the BP neural network.

The present invention will be described more specifically below with reference to preferred examples.

Preferred example 1:

the present invention will be described in further detail with reference to the accompanying drawings. The five-degree-of-freedom upper arm prosthesis control system based on the FSM comprises a multi-channel surface electromyography signal acquisition and prosthesis electromyography control system;

the multichannel surface electromyogram signal acquisition comprises a signal acquisition module and a signal processing module;

the artificial limb myoelectricity control system comprises an A/D conversion module, a BP neural network, a coding module, a motion control module, a man-machine interaction module, a training module and a storage module;

the signal acquisition module is used for sending the directly acquired unprocessed electromyographic signals to the signal processing module;

the signal processing module is used for amplifying, filtering and zeroing the received electromyographic signals to obtain recognizable electromyographic signals and sending the recognizable electromyographic signals to the A/D conversion module;

the A/D conversion module is used for converting the analog signals into digital signals and sending the digital signals to the BP neural network and the training module;

the work flow of the training mode is shown in the solid line portion of fig. 1. When a training mode is selected, the myoelectricity bracelet is worn on the forearm of a user, the user makes standard actions and gestures according to instructions, myoelectricity signals are collected and processed through multi-channel surface myoelectricity signal collection, the A/D conversion module carries out analog-to-digital conversion, the signals are input to the training module to train the BP neural network, and the trained neural network parameters are stored in the storage module.

And the storage module is used for storing the received neural network parameters.

The workflow of the usage pattern is shown in dashed lines in fig. 1. In the using stage, the BP neural network is used for reading neural network parameters from the storage module, completing initialization of the BP neural network, inputting electromyographic signals generated by a user into the BP neural network after the electromyographic signals pass through the a/D conversion module, classifying the received multi-channel electromyographic signals into a flexor longus signal L, a flexor digitalis signal S, a flexor digitorum profundus signal P, an extensor digraph signal D, a flexor radialis signal R and a synchronous signal C according to a BP neural network classification algorithm by the BP neural network, as shown in fig. 2, and sending the classification results to the coding module and the human-computer interaction module.

The coding module further classifies L, S, P, D, R with the duration exceeding the threshold T into a coding long flexor signal l, a coding superficial flexor signal s, a coding deep flexor signal p, a coding extension signal d, a coding radial wrist flexor signal r and a switching synchronous signal c according to the time sequence of the signals, and a short classification signal with the duration not exceeding the threshold T is coded into an invalid signal, so that the artificial limb keeps the original state. The first classification is to classify the signals into a long flexor signal L, a shallow flexor signal S, a deep flexor signal P, a digital extension signal D, a radial wrist flexor signal R and a synchronous signal C according to the acquired signal waveform; in order to prevent false identification, the artificial limb is classified for the second time according to the signal duration, the signals exceeding the time threshold are further classified into a coded long flexor signal l, a coded shallow flexor signal s, a coded deep flexor signal p, a coded extension signal d, a coded radial wrist flexor signal r and a switching synchronous signal c, the short classification signal with the duration not exceeding the threshold T is coded into an invalid signal, and the artificial limb is controlled by pairwise combination of secondary classification results.

Fig. 3 shows the signals l, s, p, d, r, c, which are further encoded in time series. In fig. 3, Hl represents the time from classifying into the flexor longus signal L to classifying into the flexor non-flexor longus signal, T is the judgment threshold of the signal time sequence, T is set to be 95ms, when T < Hl, the flexor longus signal L is classified as the encoded flexor longus signal L, otherwise, the flexor longus signal L is considered as the invalid signal; hs, Hp, Hd, Hr, Hc respectively represent the time from classification as superficial flexor signals S, deep flexor signals P, extensor signals D, flexor signals R and synchronizing signals C to classification as invalid signals, T is a judgment threshold of the signal time sequence, T is set to be 95ms, when T is less than Hs, Hp, Hd, Hr, Hc, the superficial flexor signals S, the deep flexor signals P, the extensor signals D, the flexor signals R and the synchronizing signals C are classified as coded superficial flexor signals S, coded deep flexor signals P, coded extensor signals D, coded radial flexor signals R and switching synchronizing signals C, otherwise, the signals are considered as invalid signals.

In the coding module, the further classification result of the electromyographic signals is sent into a coding state selection stack with the depth of 2; the coding module performs the conversion of working states in the form of a finite state machine, and comprises two working states: a wrist-hand system control state and an elbow-arm system control state; and switching between the two control states by using a switching synchronous signal c after further classification, and emptying the coding state selection stack when state switching occurs.

The coding state selection stack has 36 states, 11 states are used for controlling the state switching of the wrist-hand system and the elbow-arm system, namely, the state switching can occur when the state switching signal c appears. When the coding module is in an elbow control state, the amputation part of the upper arm and the elbow joint respectively have two motion modes, and the two motion modes respectively move and cooperate to have 8 motion states; when the coding module is in the wrist-hand control state, the hand has two motion states of loosening and grabbing, the wrist has four motion states of inward rotation, outward rotation, flexion and extension, the two motion states respectively and the coordinated motion state have 14 motion states, and the remaining 3 motion states enable the artificial limb to maintain the original state. As shown in table 1.

Table 1 transition table for motion state of artificial limb (indicates any one of signals l, s, p, d, r and c)

The control system and the motion state generated in the coding module are sent to the motion control module, and the motion control module coordinates and plans the motion of each part of the artificial limb according to different motion parts and motion directions and controls each part to move along a planned track.

And the human-computer interaction module comprises a use mode selection switch, a working state indicator light and a classification result display. The use mode selection switch is used for selecting to enter a training mode or a use mode. The working state indicator light is used for displaying whether the current working state is normal or not. The classification result display is used for displaying the classification result in the training mode so that a user can observe whether the training result is reliable or not.

As shown in fig. 4, the human-computer interaction module selects the training mode, and the user wears the myoelectric bracelet on his/her limbs to perform the operations of making a fist, relaxing, stretching the forearm, tightening the forearm, rotating the forearm, bending the wrist, and rotating the wrist in sequence to generate seven sets of standard training samples. And processing the seven groups of training samples by using a training module to generate an initialized BP neural network classifier. If the training result is not ideal, the user can choose to train again and repeat the above actions until the training result is more accurate. Then, the storage module stores the parameters of the BP neural network. After training, the user selects a use mode, the electromyographic signals subjected to A/D conversion in the mode directly enter a BP neural network without entering a training module, the electromyographic signals are firstly classified by using the network initialized in the training mode, then are secondarily classified by using a coding module, and finally a motion control module performs motion planning and controls the artificial limb to move.

It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

Preferred example 2:

the invention aims to solve the problems that the existing myoelectric artificial limb control system has less control freedom degree, unstable control effect, limited use range of a control mode based on myoelectric signal pattern recognition, difficult accurate control of a control method, inconvenient operation of the existing state conversion control mode, low state conversion accuracy and long required time, thereby providing a five-freedom upper arm artificial limb control system based on an FSM.

The invention relates to an upper arm artificial limb based on a hardware platform with five degrees of freedom, which comprises an upper arm amputation part, an elbow joint, a wrist joint and four controllable movement parts of a hand, wherein the upper arm amputation part has one degree of freedom and controls the internal rotation and the external rotation of the elbow; the elbow joint has a degree of freedom to control the flexion and extension of the elbow; the wrist joint has two degrees of freedom for controlling the internal and external rotation of the wrist and the flexion and extension of the wrist respectively; the hand has a degree of freedom that controls the release and grasping of the hand. The upper arm amputation part and the elbow joint form an elbow-arm system, and the wrist joint and the hand form a wrist-hand system.

The invention relates to a five-degree-of-freedom upper arm prosthesis control system based on an FSM (finite state machine), which comprises eight-channel surface electromyographic signal acquisition, a BP (back propagation) neural network, a coding module, a motion control module and a training module.

The eight-channel surface electromyographic signal acquisition is used for acquiring and processing electromyographic signals of a user;

and the training module is used for training the electromyographic signals acquired by standard actions of the user in a training mode to obtain the neural network parameters.

And the BP neural network is used for reading the neural network parameters, finishing the initialization of the BP neural network, classifying the electromyographic signals collected after the user acts by using the initialized BP neural network in a use mode, and sending the classification result to the coding module.

And the coding module is used for calculating the time sequence of the collected electromyographic signals and further classifying the electromyographic signals according to the duration.

The coding module is realized by adopting a state machine and has two working states: the wrist-hand system control state and the elbow-arm system control state, and the further classified myoelectric signals switch the two working states.

When the wrist and hand system is in a wrist and hand system state, the wrist and hand system is used for controlling the wrist joint and the hand according to the further classified electromyographic signals; when the elbow joint and the upper arm amputation part are controlled in the elbow and arm control state.

And the motion control module plans the motion of each joint of the artificial limb according to the control signal sent by the coding module and controls each joint to move along the planned track.

Preferably, the BP neural network classifies the received multi-channel electromyographic signals into a long flexor signal L, a superficial flexor signal S, a deep flexor signal P, a finger extension signal D, a radial wrist flexor signal R and a synchronization signal C according to a BP algorithm.

Preferably, the coding module further classifies L, S, P, D, R whose duration exceeds the threshold T into a coded long flexor signal l, a coded shallow flexor signal s, a coded deep flexor signal p, a coded extension signal d, a coded radial wrist flexor signal r and a switching synchronization signal c according to the time sequence of the signals, and a short classification signal whose duration does not exceed the threshold T is coded into an invalid signal, and the prosthesis remains in the original state.

Preferably, when the classification result of the electromyographic signals is the switching synchronization signal c, the working state of the coding module is switched, and when the state switching occurs, the coding state selection stack is emptied.

The classification result of the electromyographic signals is sent into a coding state selection stack with the depth of 2, the coding state selection stack has 36 states, 11 states are used for controlling the state switching of a wrist-hand system and an elbow-arm system, when a coding module is in the elbow-arm control state, 8 motion states exist, when the coding module is in the wrist-hand control state, 14 motion states exist, and the remaining three states enable the prosthesis to maintain the original state.

Preferably, the device comprises an A/D conversion module, and the eight-channel surface electromyographic signal acquisition comprises a signal acquisition module and a signal processing module.

And the signal acquisition module is used for sending the received unprocessed electromyographic signals to the signal processing module.

And the signal processing module is used for respectively amplifying, filtering and zeroing the received original electromyographic signals to obtain recognizable electromyographic signals and sending the recognizable electromyographic signals to the A/D conversion module.

And the A/D conversion module is used for converting the analog signals into digital signals and sending the digital signals to the BP neural network and the training module.

Preferably, the system further comprises a human-computer interaction module.

And the human-computer interaction module comprises a use mode selection switch, a working state indicator light and a classification result display.

Using a mode selection switch for selecting whether to enter a training mode;

the working state indicator light is used for displaying the current working state;

and the classification result display is used for displaying the classification result.

Preferably, the device further comprises a storage module.

And the storage module is used for receiving and storing the parameters of the BP neural network.

In the description of the present application, it is to be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like indicate orientations or positional relationships based on those shown in the drawings, and are only for convenience in describing the present application and simplifying the description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present application.

Those skilled in the art will appreciate that, in addition to implementing the systems, apparatus, and various modules thereof provided by the present invention in purely computer readable program code, the same procedures can be implemented entirely by logically programming method steps such that the systems, apparatus, and various modules thereof are provided in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system, the device and the modules thereof provided by the present invention can be considered as a hardware component, and the modules included in the system, the device and the modules thereof for implementing various programs can also be considered as structures in the hardware component; modules for performing various functions may also be considered to be both software programs for performing the methods and structures within hardware components.

The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.