CN107160419A - A kind of robot movement mechanism of electromyographic signal control - Google Patents

Abstract

The invention relates to a movement mechanism of a manipulator controlled by electromyographic signals, which comprises a main body of the manipulator (including five finger joints), a simulated wrist joint, a transmission design and a support. When the collected myoelectric signal is transmitted to the drive mechanism through the computer, through the control of three independent motors, it can realize the three types of movements of grasping and releasing of five fingers, wrist swinging up and down within a certain range, and different combinations of wrist rotation within a certain range. Among them, the grasping and releasing action of the fingers is realized by the pulling of the nylon wire driven by the motor, and the return of the return spring between the knuckles. The up and down swing of the simulated wrist joint is driven by a belt, and the rotational action of the simulated wrist joint is directly controlled by the stepper motor shaft through the key connection. The parts in the present invention are formed by 3D printing except for the support, and cooperate with the necessary mechanical standard parts, which are environmentally friendly and beautiful; at the same time, by collecting myoelectric signals and converting them into physical actions, an operation platform is provided for the research of myoelectric control. For modern medical machinery The promotion also has certain reference significance.

Description

A Manipulator Movement Mechanism Controlled by Myoelectric Signal

technical field

The invention relates to a manipulator movement mechanism controlled by electromyographic signals, which can be used in an under-actuated terminal mechanical actuator.

Background technique

With the development of science and technology, the surface electromyography system has been widely used in the medical field. Myoelectric control is to collect the movement of muscle nerves through the sensor on the surface of the skin, after processing the myoelectric signal, the movement of the transmission device is driven by the motor, so as to realize the movement of the prosthetic limb.

In order to improve the accuracy of signal processing, a graphical network based on Bayesian probability can be used, that is, Bayesian network, which is based on probabilistic reasoning and has unique advantages in solving uncertainty and incompleteness problems. applied in many fields.

In addition, the development of 3D printing technology provides a fast, beautiful, and low-cost way for parts to be formed. By combining 3D printed parts with standard mechanical parts, it can not only realize the required mechanical structure functions, but also be environmentally friendly and lightweight.

Contents of the invention

The purpose of the present invention is to provide a convenient research platform for the field of myoelectric control by converting myoelectric signals into simulated actions.

In order to achieve the above object, the technical solution of the present invention is to provide a manipulator movement mechanism controlled by myoelectric signals, which is characterized in that it includes five mechanical fingers and each mechanical finger includes a plurality of knuckles hinged from head to tail, and all knuckles are composed of Pulled by at least one wire rope, five mechanical fingers are connected to the front end of the simulated palm, and the rear end of the simulated palm is provided with a wrist joint connection shaft, which drives the simulated palm to swing up and down, and the second motor drives the wrist joint through the gear mechanism. The shaft rotates, the wrist joint connection shaft is set on the simulated wrist joint with the first motor, the first motor drives the simulated palm to turn inside and outside, the simulated wrist joint is set on the support, and five steel wire ropes pass through the support and then wind around it, which is driven by the third motor on the rotating shaft.

Preferably, a return spring is connected between two adjacent knuckles.

Preferably, three independent signal logic judgment circuits respectively judge the movements of fingers and wrists, and the three independent signal logic judgment circuits realize respectively:

The first signal logic judgment circuit judges the movement of the fingers, and sets the fingers as A1, A2, and A3, which respectively correspond to the three modes of the three actions of the motor. The signal is driven by the computer program to realize five The corresponding movement of the finger, if no movement is detected, set it to A0;

The second signal logic judging circuit judges the movement of the wrist swinging up and down, and sets the wrist to return to the initial position from the middle up, from the top, from the middle down, and from the bottom to the initial position as B1, B2, B3, and B4, corresponding to The four modes of motor two action, the signal is driven by the computer program to realize the corresponding action of the simulated palm 6, if no action is detected, set it to B0;

The third signal logic judging circuit judges the action of wrist internal and external rotation, and sets the wrist to rotate from the middle to the inside, return to the initial position from the inside, rotate from the middle to the outside, and return to the initial position from the outside as C1, C2, C3, and C4. Corresponding to the four modes of the first motor action, the signal is driven by the computer program to realize the corresponding action of the simulated wrist joint 8. If no action is detected, it is set to C0.

Preferably, according to the Bayesian network, the probability of successive occurrences of different actions under the same type of finger and wrist actions is set, that is, the probability that each independent signal logic judgment circuit controls the switching motor mode:

For the first signal logic judgment circuit, the probability of occurrence of A1, A2, and A3 is respectively 0, 0.7, and 0.3 under the condition of occurrence of A1; under the condition of occurrence of A2, the probability of occurrence of A1, A2, and A3 The probabilities are 0.5, 0, and 0.5 respectively; under the condition that A3 occurs, the probabilities of A1, A2, and A3 are 0.3, 0.7, and 0, respectively;

For the second signal logic judgment circuit, under the condition that B1 occurs, the probabilities of B1, B2, B3, and B4 are respectively 0, 1, 0, and 0; under the condition that B2 occurs, the probability of occurrence The probabilities of B1, B2, B3, and B4 are 0.5, 0, 0.5, and 0 respectively; under the condition that B3 occurs, the probabilities of B1, B2, B3, and B4 are 0, 0, 0, and 1; Under the condition that B4 occurs, the probabilities of B1, B2, B3, and B4 are 0.5, 0, 0.5, and 0, respectively;

For the third signal logic judgment circuit, under the condition that C1 occurs, the probabilities of C1, C2, C3, and C4 are respectively 0, 1, 0, and 0; under the condition that C2 occurs, the probability of occurrence The probabilities of C1, C2, C3, and C4 are 0.5, 0, 0.5, and 0 respectively; under the condition that C3 occurs, the probabilities of C1, C2, C3, and C4 are 0, 0, 0, and 1; Under the condition that C4 occurs, the probabilities of C1, C2, C3, and C4 are 0.5, 0, 0.5, and 0, respectively.

Beneficial effects of the present invention: provide a manipulator movement mechanism controlled by myoelectric signals, convert the collected myoelectric signals into circuit signals through computer programming, and then drive the mechanical transmission mechanism to realize three types of actions of the manipulator, with 3 to 4 types of actions for each type It shows that there are many combined actions between the two; combined with the Bayesian network algorithm, the association between successive actions is preset to improve the accuracy of signal recognition. The mechanism conforms to the existing mechanical structure and function, has a good appearance design, can reduce costs through 3D rapid prototyping printing technology, and builds a convenient research platform for research in the field of myoelectric control.

Description of drawings

Figure 1 is a schematic diagram of the overall shape and transmission of the manipulator;

Figure 2 is a schematic diagram of the detailed design of the mechanical finger and its related functions;

Figure 3 shows the degree of action association according to the Bayesian network setting.

detailed description

In order to make the present invention more comprehensible, preferred embodiments are described in detail below with accompanying drawings.

As shown in Fig. 1, a kind of electromyographic signal control manipulator movement mechanism provided by the present invention includes mechanical fingers 1, 2, 3, 4, 5, and there are round holes for steel wire rope 12 to pass through inside the knuckles of the mechanical fingers. There are connecting holes for five mechanical fingers on the simulated palm 6 . The wrist joint connecting shaft 7 is fixedly connected with the simulated palm 6 , and the wrist joint connecting shaft 7 is arranged on the simulated wrist joint 8 . The simulated wrist joint 8 is embedded in the circular hole of the support 9 . The transmission part of the mechanism has a transmission gear 10, which is installed on the wrist joint connecting shaft 7 and meshes with the motor transmission gear 11. The power part of the mechanism comes from three independent motors, the first motor M1, the second motor M2, and the third motor M3.

The mechanism of the present invention also includes a plurality of knuckle bolts. The knuckle connecting bolts are used to connect the knuckles of the five fingers, and the knuckles are connected by return springs 13. The circular holes inside the knuckles refer to the circular hole design of the knuckle structure in Fig. 2 . Wire rope 12 is gathered into the overhanging shaft of motor three M3 by circular hole. The transmission gear 10 is installed on the left end of the wrist joint connecting shaft 7, and is meshed with the motor transmission gear 11 on the extension shaft of the motor two M2. There is a hole for connecting the shaft on the simulated wrist joint 8, which is connected with the outstretched shaft of the motor one M1 through a key. The invention also includes a single-chip microcomputer control system.

The action part of the manipulator is the carrier for motion control, including mechanical fingers 1, 2, 3, 4, 5, simulated mechanical palm 6, wrist joint connection shaft 7, and simulated mechanical wrist 8. Five knuckles are equipped with back-moving spring 13 at joint, realize the reset after clenching. The positive and negative rotation and step motion of motor two M2 drive the gear 10 to move, thereby driving the action of the simulated palm 6; The simulated wrist joint is driven by the connecting shaft to realize internal and external rotation.

The overall mechanical structure is controlled by connecting the output signal of the computer. After the EMG signal is collected, three independent signal logic judgment circuits judge the movements of fingers and wrists respectively. The above three independent signal logic judgment circuits respectively realize:

The first signal logic judging circuit judges the movement of the fingers, and sets the grip, half grip, and openness of the fingers as A1, A2, and A3, which correspond to the three modes of motor three and M3 respectively. The signal is driven by the computer program to realize five The corresponding movement of the first finger, if no movement is detected, set it to A0;

The second signal logic judging circuit judges the movement of the wrist swinging up and down, and sets the wrist to return to the initial position from the middle up, from the top, from the middle down, and from the bottom to the initial position as B1, B2, B3, and B4, corresponding to The four modes of the motor 2 M2 action, the signal is driven by the computer program to realize the corresponding action of the simulated palm 6, if no action is detected, set it to B0;

The third signal logic judging circuit judges the action of wrist internal and external rotation, and sets the wrist to rotate from the middle to the inside, return to the initial position from the inside, rotate from the middle to the outside, and return to the initial position from the outside as C1, C2, C3, and C4. Corresponding to the four modes of motor one M1 action, the signal is driven by the computer program to realize the corresponding action of the simulated wrist joint. If no action is detected, set it to C0.

According to the Bayesian network, set the probability of successive occurrences of different actions under the same type of action, that is, the probability that each independent signal logic judgment circuit controls the switching motor mode, combined with Figure 3:

For the first signal logic judgment circuit, under the condition that A1 occurs, the probabilities of A1, A2, and A3 are respectively 0, 0.7, and 0.3; under the condition of A2, the probabilities of A1, A2, and A3 are respectively are 0.5, 0, 0.5; under the condition of A3 happening, the probabilities of A1, A2, A3 happening are 0.3, 0.7, 0 respectively;

For the second signal logic judgment circuit, under the condition that B1 occurs, the probabilities of B1, B2, B3, and B4 are respectively 0, 1, 0, and 0; The probabilities of B2, B3, and B4 are 0.5, 0, 0.5, and 0 respectively; under the condition that B3 occurs, the probabilities of B1, B2, B3, and B4 are 0, 0, 0, and 1; the default is B4 Under the conditions of occurrence, the probability of occurrence of B1, B2, B3, and B4 is 0.5, 0, 0.5, and 0, respectively;

For the third signal logic judgment circuit, the probabilities of C1, C2, C3, and C4 are respectively 0, 1, 0, and 0 under the condition that C1 occurs; The probabilities of C2, C3, and C4 are 0.5, 0, 0.5, and 0 respectively; under the condition that C3 occurs, the probabilities of C1, C2, C3, and C4 are respectively 0, 0, 0, and 1; the default is C4 Under the conditions of occurrence, the probability of occurrence of C1, C2, C3, and C4 is 0.5, 0, 0.5, and 0, respectively.

If the collected myoelectric signal is A2-B0-C3, when it is connected to the drive end, the first signal logic judgment circuit is used to drive the motor 3 M3, and the mechanical finger is pulled by the wire rope to reach the preset A2 shape, that is, half-grip , forming a shape that imitates the natural bending of the human hand; if the traction force disappears, the bent five fingers can return to the stretched state under the action of the return spring;

Through the second signal logic judgment circuit, since the preset action is not collected, the motor 2 M2 does not respond and remains in the standby state;

Through the third signal logic judgment circuit, the motor one M1 is driven, and the motor shaft steps clockwise to drive the simulated wrist to turn off and turn outward, reaching the preset C3 shape, that is, the wrist turns outward.

Claims (4)

1. a kind of robot movement mechanism of electromyographic signal control, it is characterised in that including five mechanical fingers (1,2,3,4,5), Every mechanical finger includes the finger joint that multiple head and the tail are hinged, and all finger joints are drawn by least steel cable (12), five mechanical fingers (1,2,3,4,5) are connected to the front end of simulation palm (6), and the rear end of simulation palm (6) is provided with wrist joint connecting shaft (7), and wrist is closed Saving connecting shaft (7) drives simulation palm (6) is upper and lower to swing, and motor two (M2) drives wrist joint connecting shaft via gear mechanism (7) rotate, wrist joint connecting shaft (7) is on the simulation wrist joint (8) with motor one (M1), motor one (M1) driving Simulate palm (6) inside and outside upset, simulation wrist joint (8) is located at bearing (9), five steel wire ropes (12) through after bearing (9) around In the rotary shaft for driving rotation by motor three (M3).

2. the robot movement mechanism of a kind of electromyographic signal control as claimed in claim 1, it is characterised in that two neighboring Back-moving spring (13) is connected between the finger joint.

3. the robot movement mechanism of a kind of electromyographic signal control as claimed in claim 1, it is characterised in that by three independences Action of the signal logic decision circuitry respectively to finger, wrist judge, three independent signal logic decision circuitries point Do not realize：

First signal logic decision circuitry, judges the action of finger, by finger grips, partly holds, opens and be set as A1, A2, A3, Three models of respective motor three (M3) action respectively, signal realizes five fingers by computer program-driven circuit Corresponding actions, if not detecting action, are set as A0；

Second signal logic decision circuitry, judges the action that wrist is swung up and down, and by wrist, centre is upward certainly, replied from top Initial position, certainly middle initial position setting of down, from lower section replying are B1, B2, B3, B4, difference respective motor two (M2) Four kinds of patterns of action, signal realizes the corresponding actions of simulation palm 6, if not detecting by computer program-driven circuit To action, it is set as B0；

3rd signal logic decision circuitry, judges the action of outward turning in wrist, by wrist from centre toward inward turning, at the beginning of the interior reply Beginning position, from it is middle be C1, C2, C3, C4 toward outward turning, from outer reply initial position setting, respective motor one (M1) is moved respectively The four kinds of patterns made, signal realizes the corresponding actions of simulation wrist joint, if not detecting by computer program-driven circuit To action, it is set as C0.

4. the robot movement mechanism of a kind of electromyographic signal control as claimed in claim 3, it is characterised in that according to Bayes The probability that different actions occur in succession under network, setting same class finger, list action, i.e. each independent signal logic judgment Circuit controls the probability of switch motor pattern：

For first signal logic decision circuitry, it is preset under conditions of A1 generations, occurs A1, A2, A3 probability point Wei 0,0.7,0.3；Under conditions of A2 generations, the probability for occurring A1, A2, A3 is respectively 0.5,0,0.5；The bar occurred in A3 Under part, the probability for occurring A1, A2, A3 is respectively 0.3,0.7,0；

For second signal logic decision circuitry, it is preset under conditions of B1 generations, occurs B1, B2, B3, B4 probability Respectively 0,1,0,0；It is preset under conditions of B2 generations, the probability for occurring B1, B2, B3, B4 is respectively 0.5,0,0.5,0；In advance It is located under conditions of B3 generations, the probability for occurring B1, B2, B3, B4 is respectively 0,0,0,1；It is preset under conditions of B4 generations, hair Raw B1, B2, B3, B4 probability are respectively 0.5,0,0.5,0；

For the 3rd signal logic decision circuitry, it is preset under conditions of C1 generations, occurs C1, C2, C3, C4 probability Respectively 0,1,0,0；It is preset under conditions of C2 generations, the probability for occurring C1, C2, C3, C4 is respectively 0.5,0,0.5,0；In advance It is located under conditions of C3 generations, the probability for occurring C1, C2, C3, C4 is respectively 0,0,0,1；It is preset under conditions of C4 generations, hair Raw C1, C2, C3, C4 probability are respectively 0.5,0,0.5,0.