CN112223338A - Finger control method and system for dexterous robot hand - Google Patents
Finger control method and system for dexterous robot hand Download PDFInfo
- Publication number
- CN112223338A CN112223338A CN202011020923.5A CN202011020923A CN112223338A CN 112223338 A CN112223338 A CN 112223338A CN 202011020923 A CN202011020923 A CN 202011020923A CN 112223338 A CN112223338 A CN 112223338A
- Authority
- CN
- China
- Prior art keywords
- finger
- grabbed
- current
- weight
- stroke
- 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 35
- 230000001149 cognitive effect Effects 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 7
- 239000000463 material Substances 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 239000003086 colorant Substances 0.000 claims description 3
- 238000001514 detection method Methods 0.000 abstract description 14
- 238000009530 blood pressure measurement Methods 0.000 abstract description 3
- 210000003811 finger Anatomy 0.000 description 152
- 241000220225 Malus Species 0.000 description 11
- 210000003813 thumb Anatomy 0.000 description 10
- 235000012206 bottled water Nutrition 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 235000021016 apples Nutrition 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 244000269722 Thea sinensis Species 0.000 description 2
- 238000005452 bending Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J15/00—Gripping heads and other end effectors
- B25J15/08—Gripping heads and other end effectors having finger members
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J15/00—Gripping heads and other end effectors
- B25J15/02—Gripping heads and other end effectors servo-actuated
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Manipulator (AREA)
Abstract
The invention belongs to the field of robot finger control, and discloses a finger control method and a finger control system for a dexterous robot hand, wherein the method comprises the following steps: acquiring category information of an article to be grabbed, and determining theoretical motor driving current of each finger for grabbing the article to be grabbed in a pre-established control model of the stroke and the current of a driving motor for grabbing each category of articles by each finger of a dexterous hand; adopting a theoretical motor driving current to drive each finger of the dexterous hand to grab an object to be grabbed, and acquiring an actual motor driving current of each finger for grabbing the object to be grabbed in real time; when the actual motor driving current of any finger for grabbing the object to be grabbed exceeds the theoretical motor driving current of the corresponding finger, the driving motor stops working, and the grabbing action of the object to be grabbed is completed; according to the control method, the pressure detection of the fingers is converted into the current detection of the motor for driving the fingers to move, so that the precision and convenience of pressure measurement are improved, and the precision of finger feedback control is improved; the control system has simple structure.
Description
Technical Field
The invention relates to the field of robot finger control, in particular to a method and a system for controlling fingers of a dexterous robot hand.
Background
The dexterous hand of the humanoid robot can grab different articles is an important content for the development of robot products, and because the difference of the weight and the surface hardness of the articles provides huge challenges for the adaptability and the accuracy of a finger pressure feedback control scheme. Most of the current dexterous finger pressure feedback is based on a strain gauge type pressure sensor, a strain gauge is generally pasted at the tail end of a finger, and the pressure applied to the surface of an article by the finger of a robot is detected through the strain gauge to control the movement of the finger. The strain gauge detects pressure and belongs to direct detection, and the pressure detection under the complex situation is not an ideal scheme for finger grabbing, because the strain gauge detects pressure according to multiple factors such as contact angle and contact area of the article, the scheme is difficult to realize accurate pressure feedback control, and larger errors exist.
Disclosure of Invention
Aiming at the problems in the prior art, the invention aims to provide a finger control method and a finger control system for a dexterous robot hand, wherein the method solves the problem of inaccurate finger pressure detection through a strain gauge by converting finger pressure detection into motor current detection for driving a finger to move, improves the precision and convenience of pressure measurement and greatly improves the precision of finger feedback control; the control system has simple structure and easy operation and implementation.
In order to achieve the purpose, the invention is realized by adopting the following technical scheme.
The finger control method for the dexterous robot hand comprises the following steps:
acquiring category information of an article to be grabbed; determining theoretical motor driving current of each finger of the dexterous hand for grabbing the object to be grabbed in a pre-established control model of the stroke and the current of a driving motor of each finger for grabbing each type of object of the dexterous hand according to the category information of the object to be grabbed;
driving each finger of the dexterous hand to perform grabbing action of an object to be grabbed by adopting the theoretical motor driving current, and acquiring the actual motor driving current of each finger of the dexterous hand grabbing the object to be grabbed in real time;
comparing the theoretical motor driving current of each finger of the dexterous hand grabbing the object to be grabbed with the corresponding actual motor driving current of each finger of the dexterous hand grabbing the object to be grabbed;
when the actual motor driving current of any finger grabbing the object to be grabbed exceeds the theoretical motor driving current of the corresponding finger grabbing the object to be grabbed, the driving motor stops working, and the grabbing action of the object to be grabbed is completed.
The first technical scheme of the invention has the characteristics and further improvements that:
(1) the pre-established control model for the stroke and the current of the driving motor for grabbing each type of articles by each finger of the dexterous hand comprises the following steps:
dividing the articles into different types of articles according to the weight and the hardness of the articles;
for each type of article, a relation curve of the stroke of a driving motor of each finger of the dexterous hand under the weight and the hardness of the corresponding article and the current is respectively established.
(2) The step of dividing the article into different types of articles according to the weight and hardness of the article comprises:
dividing the article into G gears according to the weight of the article, wherein the weight range of each gear is as follows:
wherein G is the total number of gears by weight, MmaxIs the maximum value by weight;
dividing the article into N gears according to the hardness of the article, wherein the hardness range of each gear is as follows:
wherein N is the total number of hardness gears, PmaxIs the maximum value of hardness;
the item is classified as a K-class item; wherein K is gxn.
(3) The step of establishing a relation curve of the stroke of the driving motor of each finger of the dexterous hand under the weight and the hardness of the corresponding article and the current for each type of article respectively comprises the following steps:
when the tail end of a single finger of the dexterous hand is hung with a load, a relation curve of the stroke of a driving motor and the current of the single finger in a 1 st weight range and a 1 st hardness range is tested;
respectively obtaining the relation curves of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the 1 st hardness range according to the relation curves of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the 1 st hardness range;
repeatedly obtaining the operation of obtaining the relation curve of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the N different hardness ranges to obtain the relation curve of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the N different hardness ranges;
and repeating the operation of obtaining the relation curve of the stroke and the current of the driving motor of the single finger under the G weight ranges and the N different hardness ranges, and obtaining the relation curve of the stroke and the current of the driving motor of each remaining finger of the dexterous hand under the G weight ranges and the N different hardness ranges.
(4) When the tail end of a single finger of the dexterous hand is hung with a load, the step of testing the relation curve of the stroke and the current of the driving motor of the single finger under the 1 st weight range and the 1 st hardness range comprises the following steps: for articles in the 1 st weight range and 1 st hardness range, the weight median in the 1 st weight range
The relation curve of the stroke of the driving motor of a single finger in the 1 st weight median and the 1 st hardness range and the current is tested.
(5) Respectively obtaining the driving power of the single finger in the 1 st weight range and the N different hardness ranges according to the relation curves of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the 1 st hardness rangeThe step of the relation curve of the machine stroke and the current comprises the following steps: the relation curve of the stroke of the drive motor and the current of a single finger in the 1 st weight median and the 1 st hardness range is translated rightwards sequentially
And (4) respectively obtaining the relation curves of the stroke and the current of the driving motor of a single finger under the 1 st weight median and N different hardness ranges by using the preset units.
(6) The step of obtaining the category information of the object to be grabbed comprises the following steps:
acquiring category characteristic information of the to-be-grabbed object through a binocular stereoscopic vision device; the category characteristic information comprises shape, size, color and material information of the article;
estimating the weight and the hardness of the object to be grabbed according to the class characteristic information of the object to be grabbed;
and determining the category information of the object to be grabbed according to the estimated weight and hardness of the object to be grabbed.
(7) The step of estimating the weight and the hardness of the object to be grabbed according to the class characteristic information of the object to be grabbed comprises the following steps of:
determining the density of the object to be grabbed according to the color and material information of the object;
calculating the volume of the object to be grabbed according to the shape and the size of the object;
and estimating the weight and the hardness of the object to be grabbed according to the density of the object to be grabbed and the volume of the object to be grabbed.
(8) The step of acquiring the category characteristic information of the object to be grabbed through the binocular stereoscopic vision device comprises the following steps of:
acquiring a first plane image and a second plane image of the article to be grabbed;
intelligently recognizing the first plane image and the second plane image to obtain specific cognitive features of the article to be grabbed, wherein the specific cognitive features comprise textures, contours and colors;
processing the first plane image and the second plane image by using a binocular vision processing algorithm to obtain a point cloud picture;
and obtaining the category characteristic information of the to-be-grabbed object according to the specific cognitive characteristics and the point cloud picture.
(II) a dexterous hand finger control system of robot, its characterized in that includes:
the binocular stereoscopic vision device is used for acquiring the class information of the to-be-grabbed object and transmitting the acquired class information of the to-be-grabbed object to the processor;
the ADC sampling circuit is arranged on each finger of the dexterous hand and is used for acquiring the actual motor driving current for grabbing the object to be grabbed by each finger of the dexterous hand in real time;
the processor comprises a storage module and a motor current determining module, wherein the storage module is used for storing a control model of the stroke and the current of a driving motor for grabbing each type of objects by each finger of the dexterous hand; the comparison module is used for comparing the theoretical motor driving current with the actual motor driving current, and sending a motor stopping signal to the finger driving motor when the actual motor driving current of any finger grabbing the object to be grabbed exceeds the theoretical motor driving current of the corresponding finger grabbing the object to be grabbed;
and the finger driving motor is arranged on each finger of the dexterous hand and is used for driving each finger of the dexterous hand to grab an article according to the theoretical motor driving current.
Compared with the prior art, the invention has the beneficial effects that:
1) in the finger control method of the robot dexterous hand, accurate measurement of finger pressure of the dexterous hand is the key for realizing flexible grabbing of an object, and the problem of inaccurate finger pressure detection through a strain gauge is solved by converting finger pressure detection into motor current detection for driving the finger to move; the control system has simple structure and easy operation and implementation.
2) The finger control method of the robot dexterous hand is not influenced by the contact angle and the contact area of the object and the tail end of the finger, and can greatly improve the precision and the convenience of finger pressure detection, thereby greatly improving the sensitivity of finger pressure feedback control of the dexterous hand.
3) The finger control method of the robot dexterous hand is suitable for soft and hard objects, the grabbing success rate is over 93 percent, wherein the hard objects can be grabbed as long as the fingers can be contacted, and the grabbing success rate of the soft objects is obviously higher than that of the existing method of passing through a strain gauge.
Drawings
The invention is described in further detail below with reference to the figures and specific embodiments.
FIG. 1 is a flow chart of a finger control method of a dexterous robot hand according to an embodiment of the present invention;
FIG. 2 is a diagram of a finger control system of a dexterous robot hand according to an embodiment of the present invention;
FIG. 3 is a diagram of an embodiment of the present invention that divides an article into four categories based on the weight and stiffness of the article;
FIG. 4 is a diagram of an embodiment of the four article types shown in FIG. 3;
FIG. 5 is a diagram illustrating the relationship between the stroke and the current of the driving motor corresponding to the thumb grasping a light and hard object according to an embodiment of the present invention;
fig. 6 is a diagram illustrating a relationship between a stroke of a driving motor and a current corresponding to a hard object gripped by a thumb according to an embodiment of the present invention.
Detailed Description
Embodiments of the present invention will be described in detail below with reference to examples, but it will be understood by those skilled in the art that the following examples are only illustrative of the present invention and should not be construed as limiting the scope of the present invention.
Referring to fig. 1, a flow chart of a finger control method of a robot dexterous hand according to an embodiment of the present invention includes the following steps:
step 1, obtaining category information of an article to be grabbed; the method comprises the following specific steps:
and acquiring the category characteristic information of the object to be grabbed through a binocular stereoscopic vision device. Specifically, a first plane image and a second plane image of an article to be grabbed are obtained; intelligently recognizing the first plane image and the second plane image to obtain specific cognitive features of the object to be grabbed, wherein the specific cognitive features comprise textures, contours and colors; processing the first plane image and the second plane image by using a binocular vision processing algorithm to obtain a point cloud picture; obtaining category characteristic information of the object to be grabbed according to the specific cognitive characteristics and the point cloud picture; the category feature information includes shape, size, color and material information of the article.
And estimating the weight and the hardness of the object to be grabbed according to the class characteristic information of the object to be grabbed. The method specifically comprises the following steps: determining the density of the object to be grabbed according to the color and material information of the object; calculating the volume of the object to be grabbed according to the shape and the size of the object; and estimating the weight and the hardness of the object to be grabbed according to the density of the object to be grabbed and the volume of the object to be grabbed.
And determining the class information of the object according to the estimated weight and hardness of the object to be grabbed. Specifically, the estimated weight and hardness of the object to be grasped are compared with the corresponding relationship between the weight and hardness of the object and the type of the object, which is established in advance, and the category information of the object to be grasped is determined.
And 2, determining theoretical motor driving current for grabbing the object to be grabbed by each finger of the dexterous hand in a pre-established control model of the stroke and the current of the driving motor for grabbing each type of object by each finger of the dexterous hand according to the type information of the object to be grabbed.
Wherein, the step of the control model of every drive motor stroke and electric current that every kind of article was grabbed to every finger of dexterous hand that sets up in advance includes:
the articles are classified into different types of articles according to their weight and hardness. The dexterous fingers need to take the weight and the hardness of an article into consideration at the same time, so that the article can be grabbed and cannot be damaged. According to the weight and the hardness of the object, a certain interval is set, and different grabbing types can be defined. According to the characteristics of the articles, a unique grabbing type is defined for each type of article, and the specific steps are as follows: dividing the article into G gears according to the weight of the article, wherein the weight range of each gear is as follows:
wherein G is the total number of gears by weight, MmaxIs the maximum value by weight;
dividing the article into N gears according to the hardness of the article, wherein the hardness range of each gear is as follows:
wherein N is the total number of hardness gears, PmaxIs the maximum value of hardness;
the item is classified as a K-class item; wherein K is gxn.
Referring to fig. 3, the articles are divided into G (G takes 2) gears according to the weight of the articles, i.e. the articles are divided into light and heavy articles. The articles are divided into N (N takes 2) gears according to the hardness of the articles, namely the articles are divided into soft articles and hard articles. The articles are classified into K-class articles, and K-gxn-4, i.e., the articles are classified into four classes of light and hard, light and soft, heavy and hard, heavy and soft articles. In practical application, the weight and the hardness can be divided into more intervals according to the use requirement so as to achieve the flexible grabbing requirement.
For each type of article, a relation curve of the stroke of a driving motor of each finger of the dexterous hand under the weight and the hardness of the corresponding article and the current is respectively established. The fingers of a dexterous robot hand are generally driven by a linear push rod motor, after the fingers touch the surface of an object, the linear push rod motor can generate locked rotor when continuously running, and under the stroke, the larger the current is, the larger the pressure generated at the tail ends of the fingers is; based on the principle, the pressure detection of the tail end of the finger can be converted into the current detection of the finger driving motor. The pressure at the tail end of the finger is related to the stroke of the finger driving motor (the stroke of the driving motor corresponds to the bending degree of the finger) and the motor current, and under the determined stroke, the finger pressure is positively related to the motor current, namely the force for grabbing the article by the finger can be controlled by controlling the current of the motor. The measurement of the pressure at the tail end of the finger is converted into the current measurement of the driving finger motor, so that the precision of pressure measurement can be greatly improved, and the precision and the flexibility of feedback control of the finger pressure are improved. The end of a single finger of a dexterous hand is hung with a load, so that the pressure generated when the end of the finger contacts the surface of an object can be simulated, and the method comprises the following specific steps:
when the tail end of a single finger of a dexterous hand is hung with a load, a relation curve of the stroke of a driving motor and the current of the single finger in a 1 st weight range and a 1 st hardness range is tested. The method specifically comprises the following steps: for articles in the 1 st weight range and 1 st hardness range, the weight median in the 1 st weight range
The relation curve of the stroke of the driving motor of a single finger in the 1 st weight median and the 1 st hardness range and the current is tested. As shown in fig. 3, under the condition that the finger tip is suspended from a load, according to the weight median corresponding to the first grabbing type (i.e. the weight range is light weight median), the relationship curve between the stroke of the driving motor and the current of a single finger under the weight median and the 1 st hardness (i.e. the hardness is hard) is tested, i.e. the relationship curve between the stroke of the driving motor and the current of the single finger under the light and hard conditions is tested.
And respectively obtaining the relation curves of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the 1 st hardness range according to the relation curves of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the 1 st hardness range. The method specifically comprises the following steps: the relation curve of the stroke of the drive motor and the current of a single finger in the 1 st weight median and the 1 st hardness range is translated rightwards sequentially
A predetermined unit, respectivelyAnd obtaining a relation curve of the stroke of the driving motor of the single finger under the 1 st weight median and N different hardness ranges. As shown in fig. 3, the obtained curve of the relationship between the stroke of the driving motor and the current of the single finger under the light and hard conditions is translated to the right by a size corresponding to the hardness (namely, the curve is translated to the right by a size corresponding to the hardness
Unit), a relation curve of the stroke of the driving motor of a single finger under light and soft conditions and the current can be obtained.
And repeating the operation of obtaining the relation curve of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the N different hardness ranges to obtain the relation curve of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the N different hardness ranges.
And repeating the operation of obtaining the relation curve of the stroke and the current of the driving motor of a single finger under G weight ranges and N different hardness ranges, and obtaining the relation curve of the stroke and the current of the driving motor of each remaining finger of the dexterous hand under G weight ranges and N different hardness ranges. As shown in fig. 3, a driving motor stroke and current relation curve of a single finger under heavy and hard conditions is tested; then, the relation curve of the stroke of the driving motor of the single finger under the heavy and hard condition and the current is translated to the right by the size with the corresponding hardness (namely, the single finger is translated to the right by the right
Unit), and obtaining a relation curve of the stroke of the driving motor of the single finger under heavy and soft conditions and the current.
And 3, driving each finger of the dexterous hand to perform grabbing action of the object to be grabbed by adopting a theoretical motor driving current, and acquiring the actual motor driving current for grabbing the object to be grabbed by each finger of the dexterous hand in real time by adopting an analog-to-digital converter (ADC) sampling circuit.
Step 4, comparing the theoretical motor driving current of each finger of the dexterous hand grabbing the object to be grabbed with the actual motor driving current of each finger of the corresponding dexterous hand grabbing the object to be grabbed; when the actual motor driving current of any finger grabbing the object to be grabbed exceeds the theoretical motor driving current of the corresponding finger grabbing the object to be grabbed, the driving motor stops working, and the grabbing action of the object to be grabbed is completed.
Specifically, when the dexterous hand grabs an article, the grabbing type is determined according to the article characteristics, and the relation between the stroke of each finger driving motor and the current is obtained according to the grabbing type. When the motor drives the fingers to move and touch the surface of an object, the current of the motor is increased along with the increase of the stroke, the actual motor driving current of each finger is continuously collected through a hardware sampling circuit (such as an ADC (analog-to-digital converter) sampling circuit, when the actual motor driving current exceeds the theoretical motor driving current of the finger for grabbing the object, the pressure generated at the tail end of the finger is enough for grabbing the object, and at the moment, the motor stops moving. When the smart hand with 5 fingers grabs an article, 5 fingers move simultaneously, the actual motor driving current of the 5 fingers detected in real time exceeds the theoretical motor driving current of the corresponding finger for grabbing the article when the actual motor driving current of the finger for grabbing the article appears for the first time, and no matter which finger is, the pressure of the finger is indicated to be enough to grab the article, the motor stops, and the grabbing action is completed.
(II) referring to figure 2, the finger control system of the robot dexterous hand of the embodiment of the present invention comprises:
the binocular stereoscopic vision device is used for acquiring the class information of the to-be-grabbed objects and transmitting the acquired class information of the to-be-grabbed objects to the processor;
the ADC sampling circuit is arranged on each finger of the dexterous hand and is used for acquiring the actual motor driving current of each finger of the dexterous hand for grabbing an object to be grabbed in real time;
the processor comprises a storage module and a motor current determining module, wherein the storage module is used for storing a control model of the stroke and the current of a driving motor for grabbing each type of objects by each finger of the dexterous hand; the comparison module is used for comparing the theoretical motor driving current with the actual motor driving current, and sending a motor stopping signal to the finger driving motor when the actual motor driving current of any finger grabbing the object to be grabbed exceeds the theoretical motor driving current of the corresponding finger grabbing the object to be grabbed;
the finger driving motor is arranged on each finger of the dexterous hand and used for driving each finger of the dexterous hand to grab an object according to the theoretical motor driving current.
The specific control method of the robot dexterous hand finger is given by taking 4 articles such as an apple, a teacup, bottled water and a mobile phone grabbed by the 5 finger dexterous hand finger as an example.
1) As shown in fig. 4, the articles were classified into 4 types with a weight of 200g and a hardness of 1mm as boundary lines: light and soft, heavy and soft, hard and light, hard and heavy, respectively corresponding to apple, bottled water, mobile phone and teacup.
2) When the thumb is loaded with 100g of weight through experimental tests, the current values corresponding to different strokes of the driving motor define the corresponding light and hard grabbing type of the stroke and current corresponding relation of the driving motor. The corresponding relation graph of the stroke and the current corresponding to the light and hard grabbing type of the thumb is shown in FIG. 5; wherein the abscissa stroke corresponds to the degree of bending of the finger.
3) And translating the curve to the right by 1mm based on a stroke and current corresponding relation graph corresponding to the light and hard grabbing type of the thumb to generate the stroke and current corresponding relation graph corresponding to the light and soft grabbing type of the thumb.
4) When the thumb is loaded with 350g of weight through experimental test, the corresponding current values of the motor in different strokes define the corresponding relationship between the strokes and the currents, and the corresponding weight and hard grabbing type are defined. The corresponding relationship graph of the stroke and the current corresponding to the heavy thumb and the hard grabbing type is shown in FIG. 6.
5) And translating the curve to the right by 1mm based on a stroke and current corresponding relation graph corresponding to the heavy and hard grabbing type of the thumb to generate the stroke and current corresponding relation graph corresponding to the heavy and soft grabbing type of the thumb.
6) And (5) repeating the step 2 to the step 5, and sequentially generating a corresponding relation graph of the lower stroke and the current of each grabbing type of the remaining four fingers.
7) Setting the object to be grabbed as an apple, determining that the grabbing type is light and soft, and obtaining theoretical motor driving currents of 5 fingers respectively grabbing the apple according to the corresponding relation graph of the strokes and the currents of the 5 fingers under the grabbing types, which is obtained in the step 2-6. The motor is controlled to drive 5 fingers to operate according to theoretical motor driving current, and actual motor driving current for grabbing the apple by each finger of the dexterous hand of the robot is obtained in real time through ADC sampling.
8) When the actual motor driving current of any finger for grabbing the apple exceeds the theoretical motor driving current of the corresponding finger for grabbing the apple, the finger pressure is proved to meet the grabbing condition, the driving motor stops working, namely all the fingers stop moving, and the grabbing operation of the apple is completed.
9) And 7, repeating the steps 7-8, and sequentially finishing grabbing of the teacup, the bottled water and the mobile phone.
Test of
The control sensitivity of the robot dexterous finger control method based on current detection and the existing method for detecting finger pressure through the strain gauge type pressure sensor is tested, under the same test condition, 4 objects, namely apples, tea cups, bottled water and mobile phones, are continuously grabbed for 100 times, and the grabbing success times are shown in table 1.
TABLE 1
As can be seen from table 1, compared with the existing method for detecting finger pressure by using a strain gauge type pressure sensor, the skillful robot finger control method provided by the embodiment of the invention has a high precision success rate for grabbing various articles such as apples, tea cups, bottled water and mobile phones, the grabbing success rate is up to more than 93%, and especially the grabbing success rate for soft articles such as apples and bottled water is obviously higher than that of the existing method for detecting finger pressure by using a strain gauge type pressure sensor.
Although the present invention has been described in detail in this specification with reference to specific embodiments and illustrative embodiments, it will be apparent to those skilled in the art that modifications and improvements can be made thereto based on the present invention. Accordingly, such modifications and improvements are intended to be within the scope of the invention as claimed.
Claims (10)
1. A finger control method for a dexterous robot hand is characterized by comprising the following steps:
acquiring category information of an article to be grabbed;
determining theoretical motor driving current of each finger of the dexterous hand for grabbing the object to be grabbed in a pre-established control model of the stroke and the current of a driving motor of each finger for grabbing each type of object of the dexterous hand according to the category information of the object to be grabbed;
driving each finger of the dexterous hand to perform grabbing action of an object to be grabbed by adopting the theoretical motor driving current, and acquiring the actual motor driving current of each finger of the dexterous hand grabbing the object to be grabbed in real time;
comparing the theoretical motor driving current of each finger of the dexterous hand grabbing the object to be grabbed with the corresponding actual motor driving current of each finger of the dexterous hand grabbing the object to be grabbed;
when the actual motor driving current of any finger grabbing the object to be grabbed exceeds the theoretical motor driving current of the corresponding finger grabbing the object to be grabbed, the driving motor stops working, and the grabbing action of the object to be grabbed is completed.
2. A method of controlling the finger of a robotic dexterous hand according to claim 1, wherein said pre-established control model of drive motor stroke and current for each finger gripping of each type of article of the dexterous hand comprises:
dividing the articles into different types of articles according to the weight and the hardness of the articles;
for each type of article, a relation curve of the stroke of a driving motor of each finger of the dexterous hand under the weight and the hardness of the corresponding article and the current is respectively established.
3. The method of controlling the fingers of a robotic dexterous hand of claim 2, wherein said step of dividing the objects into different categories of objects according to weight and stiffness of the objects comprises:
dividing the article into G gears according to the weight of the article, wherein the weight range of each gear is as follows:
wherein G is the total number of gears by weight, MmaxIs the maximum value by weight;
dividing the article into N gears according to the hardness of the article, wherein the hardness range of each gear is as follows:
wherein N is the total number of hardness gears, PmaxIs the maximum value of hardness;
the item is classified as a K-class item; wherein K is gxn.
4. A method of controlling the fingers of a robotic dexterous hand according to claim 3, wherein said step of establishing a drive motor stroke versus current curve for each finger of the dexterous hand at a corresponding article weight and stiffness, respectively, for each type of article, comprises:
when the tail end of a single finger of the dexterous hand is hung with a load, a relation curve of the stroke of a driving motor and the current of the single finger in a 1 st weight range and a 1 st hardness range is tested;
respectively obtaining the relation curves of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the 1 st hardness range according to the relation curves of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the 1 st hardness range;
repeatedly obtaining the operation of obtaining the relation curve of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the N different hardness ranges to obtain the relation curve of the stroke and the current of the driving motor of the single finger in the 1 st weight range and the N different hardness ranges;
and repeating the operation of obtaining the relation curve of the stroke and the current of the driving motor of the single finger under the G weight ranges and the N different hardness ranges, and obtaining the relation curve of the stroke and the current of the driving motor of each remaining finger of the dexterous hand under the G weight ranges and the N different hardness ranges.
5. The method of claim 4, wherein the step of testing the relationship between the stroke of the drive motor and the current for a single finger in the 1 st weight range and the 1 st stiffness range while suspending a load from the end of the single finger of the dexterous hand comprises: for articles in the 1 st weight range and 1 st hardness range, the weight median in the 1 st weight range
And testing the relation curve of the stroke of the driving motor of the single finger under the 1 st weight median and the 1 st hardness range and the current.
6. The method of claim 5, wherein the step of obtaining the relationship between the stroke and the current of the driving motor of the single finger in the 1 st weight range and the N different hardness ranges according to the relationship between the stroke and the current of the driving motor of the single finger in the 1 st weight range and the 1 st hardness range comprises: sequentially translating the relation curve of the stroke of the drive motor and the current of the single finger to the right in the range of the 1 st weight median and the 1 st hardness
A preset listAnd respectively obtaining the relation curves of the stroke and the current of the driving motor of the single finger under the 1 st weight median and N different hardness ranges.
7. The method of controlling the fingers of a dexterous robot hand of claim 2, wherein said step of obtaining information on the type of article to be grasped comprises:
acquiring category characteristic information of the to-be-grabbed object through a binocular stereoscopic vision device; the category characteristic information comprises shape, size, color and material information of the article;
estimating the weight and the hardness of the object to be grabbed according to the class characteristic information of the object to be grabbed;
and determining the category information of the object to be grabbed according to the estimated weight and hardness of the object to be grabbed.
8. The method of controlling fingers of a dexterous robot hand according to claim 7, wherein the step of estimating the weight and stiffness of the object to be gripped based on the class characteristic information of the object to be gripped comprises:
determining the density of the object to be grabbed according to the color and material information of the object;
calculating the volume of the object to be grabbed according to the shape and the size of the object;
and estimating the weight and the hardness of the object to be grabbed according to the density of the object to be grabbed and the volume of the object to be grabbed.
9. The finger control method of a robot dexterous hand according to claim 7, wherein the step of acquiring the category characteristic information of the object to be grabbed through a binocular stereo vision device comprises:
acquiring a first plane image and a second plane image of the article to be grabbed;
intelligently recognizing the first plane image and the second plane image to obtain specific cognitive features of the article to be grabbed, wherein the specific cognitive features comprise textures, contours and colors;
processing the first plane image and the second plane image by using a binocular vision processing algorithm to obtain a point cloud picture;
and obtaining the category characteristic information of the to-be-grabbed object according to the specific cognitive characteristics and the point cloud picture.
10. A finger control system for a robotic dexterous hand, comprising:
the binocular stereoscopic vision device is used for acquiring the class information of the to-be-grabbed object and transmitting the acquired class information of the to-be-grabbed object to the processor;
the ADC sampling circuit is arranged on each finger of the dexterous hand and is used for acquiring the actual motor driving current for grabbing the object to be grabbed by each finger of the dexterous hand in real time;
the processor comprises a storage module and a motor current determining module, wherein the storage module is used for storing a control model of the stroke and the current of a driving motor for grabbing each type of objects by each finger of the dexterous hand; the comparison module is used for comparing the theoretical motor driving current with the actual motor driving current, and sending a motor stopping signal to the finger driving motor when the actual motor driving current of any finger grabbing the object to be grabbed exceeds the theoretical motor driving current of the corresponding finger grabbing the object to be grabbed;
and the finger driving motor is arranged on each finger of the dexterous hand and is used for driving each finger of the dexterous hand to grab an article according to the theoretical motor driving current.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011020923.5A CN112223338B (en) | 2020-09-25 | 2020-09-25 | Robot smart hand finger control method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011020923.5A CN112223338B (en) | 2020-09-25 | 2020-09-25 | Robot smart hand finger control method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112223338A true CN112223338A (en) | 2021-01-15 |
| CN112223338B CN112223338B (en) | 2023-12-29 |
Family
ID=74107678
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011020923.5A Active CN112223338B (en) | 2020-09-25 | 2020-09-25 | Robot smart hand finger control method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112223338B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112884825A (en) * | 2021-03-19 | 2021-06-01 | 清华大学 | Deep learning model-based grabbing method and device |
| CN116766212A (en) * | 2023-08-17 | 2023-09-19 | 深圳市心流科技有限公司 | Bionic hand control method, bionic hand control device, bionic hand control equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105751196A (en) * | 2016-04-12 | 2016-07-13 | 华南理工大学 | Operating method on basis of master-slave industrial robot collaboration |
| US20190321984A1 (en) * | 2018-04-24 | 2019-10-24 | Fanuc Corporation | Robot controller and system |
| CN110733053A (en) * | 2019-11-04 | 2020-01-31 | 比率机器人科技(北京)有限公司 | clamping jaw for automatic catering device and clamping method thereof |
| WO2020044868A1 (en) * | 2018-08-31 | 2020-03-05 | 豊田合成株式会社 | Robot hand |
| CN111168710A (en) * | 2020-03-06 | 2020-05-19 | 安徽大学 | Dexterous finger surface pressure control method based on control current |
-
2020
- 2020-09-25 CN CN202011020923.5A patent/CN112223338B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105751196A (en) * | 2016-04-12 | 2016-07-13 | 华南理工大学 | Operating method on basis of master-slave industrial robot collaboration |
| US20190321984A1 (en) * | 2018-04-24 | 2019-10-24 | Fanuc Corporation | Robot controller and system |
| WO2020044868A1 (en) * | 2018-08-31 | 2020-03-05 | 豊田合成株式会社 | Robot hand |
| CN110733053A (en) * | 2019-11-04 | 2020-01-31 | 比率机器人科技(北京)有限公司 | clamping jaw for automatic catering device and clamping method thereof |
| CN111168710A (en) * | 2020-03-06 | 2020-05-19 | 安徽大学 | Dexterous finger surface pressure control method based on control current |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112884825A (en) * | 2021-03-19 | 2021-06-01 | 清华大学 | Deep learning model-based grabbing method and device |
| CN112884825B (en) * | 2021-03-19 | 2022-11-04 | 清华大学 | Deep learning model-based grabbing method and device |
| CN116766212A (en) * | 2023-08-17 | 2023-09-19 | 深圳市心流科技有限公司 | Bionic hand control method, bionic hand control device, bionic hand control equipment and storage medium |
| CN116766212B (en) * | 2023-08-17 | 2023-12-19 | 深圳市心流科技有限公司 | Bionic hand control method, bionic hand control device, bionic hand control equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112223338B (en) | 2023-12-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112223338A (en) | Finger control method and system for dexterous robot hand | |
| CN108656147B (en) | Self-adaptive mechanical gripper | |
| US10059007B1 (en) | Finger-aligned extendable gripping mechanisms for item manipulation | |
| CN110046538A (en) | Grip device, determines method, learning device, model and method at grasping system | |
| CN113733149B (en) | Soft finger manipulator for sensing fruit hardness and detection method | |
| Schöpfer et al. | Using a piezo-resistive tactile sensor for detection of incipient slippage | |
| CN113618709B (en) | Multi-mode force control nondestructive grabbing device for intelligent production line | |
| Taunyazov et al. | Extended tactile perception: Vibration sensing through tools and grasped objects | |
| Song et al. | Sensing shear forces during food manipulation: Resolving the trade-off between range and sensitivity | |
| CN110757463A (en) | Manipulator grabbing force control method and device | |
| CN206132939U (en) | Testing arrangement of mechanical key release button in automotive electronics key | |
| CN117428775A (en) | Force control method, device, equipment and medium for tail end clamping jaw of robot | |
| Zhang et al. | Slip detection by array-type pressure sensor for a grasp task | |
| CN209396102U (en) | A kind of soft touching grabbing device of type self-adapting type article | |
| Li et al. | Learning to detect slip for stable grasping | |
| CN116678307A (en) | Curvature measurement system and method for flexible mechanical claw | |
| CN110722591A (en) | Force sensing method for flexible body, gripping jaw, and manipulator | |
| CN216154951U (en) | Commodity circulation is arm for letter sorting with recognition function | |
| Yang et al. | A new slip-detection method based on pairwise high frequency components of capacitive sensor signals | |
| CN211310170U (en) | Coiled material manipulator stacking gripper | |
| CN112060123A (en) | Flat-clamping touch mechanical claw device and using method | |
| Wu et al. | An automatic grasp system with sensor feedback based on soft gripper | |
| CN213034649U (en) | Flat-clamping touch mechanical claw device | |
| Ayral et al. | Spectro-Temporal Recurrent Neural Network for Robotic Slip Detection with Piezoelectric Tactile Sensor | |
| Matsuno et al. | Strategy of picking up thin plate by robot hand using deformation of soft fingertip |
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 |