CN107225576B - A kind of control method based on soft finger - Google Patents
A kind of control method based on soft finger Download PDFInfo
- Publication number
- CN107225576B CN107225576B CN201710637965.5A CN201710637965A CN107225576B CN 107225576 B CN107225576 B CN 107225576B CN 201710637965 A CN201710637965 A CN 201710637965A CN 107225576 B CN107225576 B CN 107225576B
- Authority
- CN
- China
- Prior art keywords
- target object
- finger
- soft
- soft finger
- force
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 230000033001 locomotion Effects 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims description 27
- 230000006870 function Effects 0.000 claims description 19
- 238000012795 verification Methods 0.000 claims description 16
- 239000000463 material Substances 0.000 claims description 8
- 230000003190 augmentative effect Effects 0.000 claims description 4
- 230000005484 gravity Effects 0.000 claims description 4
- 229920002379 silicone rubber Polymers 0.000 claims description 4
- 239000004945 silicone rubber Substances 0.000 claims description 4
- 230000003068 static effect Effects 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000005253 cladding Methods 0.000 claims description 3
- 238000006073 displacement reaction Methods 0.000 claims description 3
- 230000005489 elastic deformation Effects 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 210000001145 finger joint Anatomy 0.000 claims 1
- 230000003993 interaction Effects 0.000 claims 1
- 238000004519 manufacturing process Methods 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 3
- 230000003416 augmentation Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 239000011521 glass Substances 0.000 description 2
- 230000035807 sensation Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000003292 glue Substances 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1628—Programme controls characterised by the control loop
- B25J9/1633—Programme controls characterised by the control loop compliant, force, torque control, e.g. combined with position control
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Manipulator (AREA)
Abstract
The present invention relates to a kind of control methods based on soft finger, which comprises step 1: robot determines the target object for needing to grab；Step 2: obtaining the shape of determining target object；Robot obtains the shape of target object by modes such as image sensing or infrared reflections；Step 3: robot determines the position of target object, and drives soft finger grip target object；Step 4: during grabbing target object and subsequent movement, minimum grip is calculated in real time, and the clamping of soft finger is carried out based on minimum grip, soft finger and the contact area being crawled between object can be increased to the full extent, grasp force is reduced, reduce impact force；Soft finger model is introduced simultaneously, solves grasping force control, bring computation complexity when can reduce multiple fingers while accommodating, more particularly to calculate moment chucking power in the control of real-time power to carry out finger control.
Description
[technical field]
The present invention relates to robot fields, and in particular to a kind of control method based on soft finger.
[background technique]
Manipulator is the certain holding functions that can imitate manpower and arm, to by fixed routine crawl, carrying object or behaviour
Make the automatic pilot of tool.In order to object be caught by not damaged due to chucking power is excessive to object, to solve
Certainly two critical issues, first is that pressure sensation and the sliding robot touch sensor for feeling function can be perceived and be measured；Second is that should apply more
Big chucking power.
During stablizing crawl and flexibly using tool, the flexibility of human finger plays critically important effect.
With the fast development of new technology, the application of anthropomorphic robot is also more and more extensive.The finger of manipulator in the prior art is big
Multi-purpose rigid material is made.Rigid finger and the contact area being crawled between article are small, and required grip is also very big, no
It is suitble to article (such as glass) made of crawl brittleness, friable material.In addition, minimum with rigid finger grip target object
Three fingers are needed just to be able to achieve stable crawl.Based on above-mentioned problems, need now a kind of new based on soft finger
Control method can be increased soft finger and the contact area being crawled between object to the full extent, be subtracted using the method
Small grasp force reduces impact force；Soft finger model is introduced simultaneously, solves grasping force control, it is same to can reduce multiple fingers
When bring computation complexity when accommodating, more particularly to calculate moment chucking power in the control of real-time power to carry out finger
Control.Especially solves the minimum grip algorithmic issue based on the Apery manipulator with multiple soft fingers.
[summary of the invention]
In order to solve the above problem in the prior art, The technical solution adopted by the invention is as follows: a kind of be based on soft hand
The control method of finger, described method includes following steps:
Step 1: robot determines the target object for needing to grab；Specifically: robot task based access control is assigned, and search needs
The target object to be selected of crawl；The verification that target object to be selected is carried out after finding target object to be selected, by by verification to
Target object is selected to be determined as target object；
Robot task based access control is assigned, and search needs the target object to be selected grabbed；Specifically: robot passes through man-machine friendship
Mutual interface or remote terminal reception task assignment, parse the task of appointment, and when task is crawl task, determination needs to grab
Target object and its essential attribute information and depth attribute information, based on to be selected in essential attribute information search working space
Target object；Wherein, essential attribute information describes the most basic attribute of target object, and the essential attribute is to be easy to obtain
Attribute information；
The verification that target object to be selected is carried out after finding target object to be selected, the target object to be selected by verification is true
It is set to target object；Specifically: the verification of target object is carried out based on depth attribute information；Wherein, depth attribute information describes
The detailed attribution information other than primary attribute of target object；The depth attribute further calculated or
Start one or more sensor to obtain；
Step 2: obtaining the shape of determining target object；The shape is used in the calculating of minimum grip；
Step 3: robot determines the position of target object, and drives soft finger grip target object；
Step 4: during grabbing target object and subsequent movement, calculating minimum grip in real time, and grab based on minimum
Power carries out the clamping of soft finger.
Further, step 4 specifically: by the essential attribute and depth attribute information of target object, determine object
Whether body is brittle body, if so, calculating minimum grip in real time, and the clamping of soft finger is carried out based on minimum grip；
The minimum grip is finger tip and object keeps press contacts and the opposite the smallest grip slided does not occur；If target object is not
It is brittle body, does not then calculate minimum grip, the quality based on object searches default grip, and is carried out based on the default grip
The clamping of target object.
Further, the minimum grip of the real-time calculating in step 4, including, step A1: establish the linear contact of soft finger
Model；Step A2: manipulator target object crawl model is established；A3: the model established based on step A1 and A2, it is bright by glug
Day multiplier calculates multiple soft fingers to the minimum grip of object.
Further, the step A1: establishing the linear contact model of soft finger, specifically: each soft hand of setting
Linear relationship between the contact force F and finger deformation quantity d of finger；
Wherein: formula (1) uses the central point of each soft finger top surface for LT local coordinate system (the n finger of origin
There is n LT local coordinate system)；Contact force of the F between soft finger and target object, n soft fingers are n contact point
Contact force is expressed as f1~fn；E is preset value, and the preset value is related to the material of finger, can rule of thumb or convergence
Algorithm determines；D is the deformation quantity maximum value of soft finger in the z-direction；It is volume point, indicates soft finger and target object
Lap, the lap because contact force cause elastic deformation and formed；K is setting constant；BC is with any
The angle [alpha] deformation quantity of soft finger in the z-direction when contacting；String when r is soft finger maximum distortion；X1, x2 are in the x-direction
Range of integration；R is the radius of soft finger；X, y are x/y plane coordinates.
Further, the step A2: establishing manipulator target object crawl model, specifically: A21 establishes manipulator and grabs
Modulus type and the soft finger contact model of A22.
Further, A21: establishing manipulator crawl model, specifically:
The equation of motion of target object is established, the equation of motion of target object is using coordinate system 0；
Wherein；
Wherein: fi is contact force namely i-th finger contact target object of i-th of soft finger i-th of contact point
Generated fix, fiy, the resultant force of tri- power of fiz, wherein fix, fiz are the tangential component of fi respectively, and fiy is the normal direction of fi
Component；Fc is the contact force of multiple soft fingers for being indicated with the form of vector to target object,It is target object
Inertial matrix,It is the position and direction of object,It is Coriolis item,It is gravity item,
It is spanner matrix, depending on the posture of position and object with contact point, is able to carry out the conversion of coordinate system.
Further, the A22: establishing soft finger contact model, specifically:
fc=W#Fr+ Ny formula (3)
Multiple soft fingers are divided into two parts, W to the contact force of object#FrReferred to as operating force, Ny (i.e. N × y) are to grasp
When the internal force that generates, internal force do not generate displacement；
It is further:
W#Fr=[fo1 T fo2 T … fon T]T
N=[N1 T N2 T … Nn T]T
Wherein:It is the pseudo inverse matrix of W,It is the resultant force acted on target object, uses
Target object coordinate system 0, N are the null space matrix of W, and the base of the kernel of W isColumn matrix formed
,It is any vector,It is the internal force generated when grasping；foiIt is that lower i-th of the finger of coordinate system 0 connects
Touch contact force caused by object；NiIt is the N matrix of i-th of finger；Pseudo inverse matrix is the generalized form of inverse matrix；
In contact point PiInternal unit normal vector can indicate are as follows:
Wherein；eNiThe internal unit normal vector for being i-th of finger in normal direction, gradS (Pi) be contact point Pi gradient,
S (x, y, z)=0 is the shape of target object；
Constraint condition g is seti(y): giIt (y) is the constraint condition for being applied to i-th finger；
gi(y)≤0 for i=1 ..., 2n formula (5)
It is specific:
Wherein；eNi-nAnd foi-nSubscript i value range be i=n+1 ..., 2n, μiIt is static friction coefficient, μsiIt is to turn round
Moment coefficient, E3 are the preset values based on soft finger material and finger structure, and can be in manipulator in crawl target object
It is modified value in the process；N is the quantity of soft finger；B0 and Bi (B1~Bn) is intermediate variable.
Further, soft finger has there are three independent joint, and finger tip is hemispherical, one layer of silicone rubber of cladding outside finger tip
Glue.
Further, it in manipulator target object crawl model, establishes using the mass center of target object as the original of coordinate system
The coordinate system 0 of point is target object；P is contact point, and P1~Pn is the contact point of n soft fingers, is established with soft finger and mesh
The contact point Pi for marking object is the local coordinate system Li of origin.
Further, the A3: the model established based on step A1 and A2 is calculated multiple soft by Lagrange multiplier
Finger to the minimum grip of object, specifically:
It is equivalent by contact force fCThe setting of minimum are as follows:For seeking the minimum value of h (y), wherein
Y is vector variable；
H (y)=(W is set#Fr+Ny)T(W#Fr+Ny)；
It introduces a Lagrange multiplier λ and carries out constrained optimization, constraint condition function g (y) and original function h (y) is contacted
To the equation equation equal with variable quantity together, is made into, to find a function minimum of the h (y) under the constraint condition of g (y)=0
Value；Lagrange multiplier λ is indicated are as follows:Obtain the augmentation Lagrange of minimization problem
Function L (y；λ):
L(y；λ)=h (y)+λTG (y) formula (6)
Wherein；L(y；λ) the Augmented Lagrangian Functions to solve minimum value；
The necessary condition for generating the local minimum of h (y) is:
It is obtained from above-mentioned formula:
Obtain y when minimum value h (y)oValue after, be arranged y=yo, the smallest internal force Ny is calculated, and then by formula (3)
Minimal-contact power fc is obtained as minimum grip.
Using the present invention, soft finger and the contact area being crawled between object can be increased to the full extent, reduced
Grasp force reduces impact force；Soft finger model is introduced simultaneously, is solved grasping force control, be can reduce multiple fingers simultaneously
Bring computation complexity when aid, more particularly to calculate moment chucking power in the control of real-time power to carry out finger control
System.
[Detailed description of the invention]
Described herein the drawings are intended to provide a further understanding of the invention, constitutes part of this application, but
It does not constitute improper limitations of the present invention, in the accompanying drawings:
Fig. 1 is the schematic diagram of soft finger deformation model of the invention.
Fig. 2 is the schematic diagram of manipulator crawl model of the invention.
Fig. 3 is the schematic diagram of soft finger contact model of the invention.
[specific embodiment]
Come that the present invention will be described in detail below in conjunction with attached drawing and specific embodiment, illustrative examples therein and says
It is bright to be only used to explain the present invention but not as a limitation of the invention.
It is a kind of control method based on soft finger applied by the present invention, the method includes as follows referring to attached drawing 1
Step:
Step 1: robot determines the target object for needing to grab；Specifically: robot task based access control is assigned, and search needs
The target object to be selected of crawl；The verification that target object to be selected is carried out after finding target object to be selected, by by verification to
Target object is selected to be determined as target object；
Robot task based access control is assigned, and search needs the target object to be selected grabbed；Specifically: robot passes through man-machine friendship
Mutual interface or remote terminal reception task assignment, parse the task of appointment, and when task is crawl task, determination needs to grab
Target object and its essential attribute information and depth attribute information, based on to be selected in essential attribute information search working space
Target object；Wherein, essential attribute information describes the most basic attribute of target object, and the essential attribute is to be easy to obtain
Attribute information；
Such as: it is specified in crawl task and needs to grab cup, then the task is crawl task, and determining needs in the task
The target object to be grabbed is cup, and attribute information is white, glass, cup；Based on the attribute information, in working space
The target object of search attribute information matches；
The verification that target object to be selected is carried out after finding target object to be selected, the target object to be selected by verification is true
It is set to target object；Specifically: the verification of target object is carried out based on depth attribute information；Wherein, depth attribute information describes
The detailed attribution information other than primary attribute of target object；Preferred: depth attribute information includes: size, light transmission
The information such as degree, volume, the depth attribute are further calculated or are started one or more sensor to obtain；
Such as: the depth attribute information of target object to be selected is obtained by sensor, carries out all targets to be selected one by one
The verification of object；Target object to be selected for one carries out the comparison item by item of its depth attribute, when all depth attributes match
When, it is determined that one target to be selected is determining target object；
Step 2: obtaining the shape of determining target object, the shape is used in the calculating of minimum grip；
Preferred: robot obtains the shape of target object by modes such as image sensing or infrared reflections, by shape
Shape is expressed as S (x, y, z)=0；(x, y, z) is space coordinate；
Step 3: robot determines the position of target object, and drives soft finger grip target object；
Preferably, it is for two or multiple to grab the fingers number of target object.
Step 4: during grabbing target object and subsequent movement, calculating minimum grip in real time, and grab based on minimum
Power carries out the clamping of soft finger；
It is preferred: by the essential attribute information and depth attribute information of target object, to determine whether target object is crisp
Property object, if so, calculating minimum grip in real time, and the clamping of soft finger is carried out based on minimum grip；If object
Body is not brittle body, then does not calculate minimum grip, and the quality based on object searches default grip, and is based on the default grip
Carry out the clamping of target object；Save the corresponding relationship between quality and default grip；The quality passes through volume and densitometer
It obtains；
The minimum grip is finger tip and object keeps press contacts and the opposite the smallest grip slided does not occur, without meter
Calculate the chucking power of each moment；Present method solves the minimum grip calculations based on the Apery manipulator with multiple soft fingers
Method problem；
It is preferred: will to be used to perceive and measure pressure sensation and the sliding robot touch sensor for feeling function is mounted on the recessed of finger
In slot, accurately to perceive and measure the size of the distance between finger and target object and grip.
It is preferred: to clamp the weight of the size of required grip and soft or hard, clamping the position of mechanical finger, target object
The factors such as inertia force, vibration and transmission efficiency that the target object measure, being crawled generates during exercise are related, in crawl and movement
Minimum grip is calculated during target object in real time；
The minimum grip of real-time calculating in step 4, specifically:
Step A: two models are established, the first model: the linear contact model of soft finger, the second model: manipulator mesh
Mark grasping body model；
A1: establishing the linear contact model of soft finger, specifically: the contact force F and finger of each soft finger of setting
Linear relationship between deformation quantity d；
Wherein: formula (1) uses the central point of each soft finger top surface for LT local coordinate system (the n finger of origin
There are n corresponding LT local coordinate systems)；Contact force of the F between soft finger and target object, n soft fingers connect at n
The contact force of contact is expressed as f1~fn；E is preset value, and the preset value is related to the material of finger, can rule of thumb or
Person convergence algorithm determines；D is the deformation quantity maximum value of soft finger in the z-direction；It is volume point, indicates soft finger and mesh
The lap of object is marked, the lap is formed because of the elastic deformation of contact force initiation；K is setting constant；BC is
The deformation quantity of soft finger in the z-direction when being contacted with any angle α；String when r is soft finger maximum distortion；X1, x2 are along x
The range of integration in direction is the lap upper minimum value and maximum value changed in the x-direction；R is the radius of soft finger；x,
Y is the x/y plane coordinate of the linear contact model of soft finger；
Attached drawing 1 is the sectional view of Y=0,It is the z value of arbitrary point C (x, y, z) on spherical surface, (attached
It is the value of AC in Fig. 1).BC be when being contacted with any angle α soft finger deformation quantity in the z-direction namely BC be soft finger with
The value of overlapped object part in the z-direction is absolute decrement of the soft finger in point C,It is opposite compress
Amount.
A2: establishing manipulator target object crawl model, specifically: establish manipulator crawl model and the contact of soft finger
Model；
Manipulator target object grabs model and is made of the target object and multi-finger manipulator for being in three-dimensional space；Manipulator
It is made of (n >=2) n soft fingers, and by the finger tip contacts target object of soft finger；Soft finger is contacted, finger tip
To object in addition to applying contact force, also apply a torque along normal direction；
Preferred: soft finger is made of silicone rubber；It considerably increases mechanical finger and is crawled connecing between object
Contacting surface product, reduces grasp force；Preferred: soft finger has there are three independent joint, and finger tip is hemispherical, cladding outside finger tip
One layer of silicone rubber；
In manipulator crawl model, establishing by the coordinate system 0 of the origin of coordinate system of the mass center of target object is target
Object, P are contact point (contact points that P1~Pn is n soft fingers)；It establishes with the contact point Pi of soft finger and target object
For the local coordinate system Li of origin.
It is preferred: force-closed model to be established during grasping body, force-closed model ensures that any external force can obtain
Balance；Force-closed to ensure that any external force be balanced, force-closed is the premise for stablizing crawl.Based on this, minimum is being carried out
It needs to meet finger in the calculating process of grip to contact with each other with target object and fricton-tight, the contact force fi at each contact point
It must be positioned in cone of friction, i.e., the tangential component of contact force must be not more than the product of its normal component and coefficient of friction；
In order to guarantee that multifinger hand can complete stable crawl, during operation, the crawl of multifinger hand be must satisfy
The condition of dynamic balance, i.e., all power (grasp force and external force including finger) its vector sums acted on target object are necessary for
Zero, if the power that finger acts on contact point can claim the crawl to meet with any force screw of the negative function on object
Force-closed condition.It usually requires that in the operating process of crawl object and meets force-closed condition.If crawl meets force-closed condition,
The grasp force of each finger can be solved according to the external force acted on object.Any power system acted on rigid body is ok
The concentrated force along certain straight line and the torque around the straight line are synthesized, this generalized force is force screw；
A21: establishing manipulator crawl model, specifically:
The equation of motion of target object is established, the equation of motion of target object is using coordinate system 0；
Wherein；
Wherein: fi is contact force namely i-th finger contact target object of i-th of soft finger i-th of contact point
Generated fix, fiy, the resultant force of tri- power of fiz, wherein fix, fiz are the tangential component of fi respectively, and fiy is the normal direction of fi
Component；Fc is the contact force of multiple soft fingers for being indicated with the form of vector to target object,It is target object
Inertial matrix,It is the position and direction of object,It is Coriolis item,It is gravity item,
It is spanner matrix, depending on the posture of position and object with contact point, is able to carry out the conversion of coordinate system；
A22: establishing soft finger contact model, specifically:
fc=W#Fr+ Ny formula (3)
Multiple soft fingers are divided into two parts, W to the contact force of object#FrReferred to as operating force, Ny (i.e. N × y) are to grasp
When the internal force that generates, internal force do not generate displacement；It is preferred: internal force refer to its acted on when grasping, at contact point target object but
The power that target plays the total stress and moment variations of object is not drawn, internal force makes finger and object meet static friction constraint, prevents edge
Body surface sliding；
It is further:
W#Fr=[fo1 T fo2 T … fon T]T
N=[N1 T N2 T … Nn T]T
Wherein:It is the pseudo inverse matrix of W,It is the resultant force ((resultant force acted on target object
Consider gravity, Coriolis force, inertia force；Using target object coordinate system 0), N is the null space matrix of W, and the zero of W is empty
Between base beColumn matrix formed,It is any vector,It is when grasping
The internal force of generation；foiIt is contact force caused by lower i-th of the finger of coordinate system 0 contact object,；NiIt is the N square of i-th of finger
Battle array；Pseudo inverse matrix is the generalized form of inverse matrix；
The null space matrix of A is one group by ker (A)={ x ∈ Cn: Ax=0 } n-dimensional vector that defines, i.e. system of linear equations
The set of all solution x of Ax=0；Wherein A is the matrix of a m*n.Matrix can regard one group of column vector C1, C2 ..., Cn as,
If this group of vector is linear independence, the solution space of AX=0 only includes vector: null vector., whereas if zero is empty
Between include non-vanishing vector, illustrate that matrix column SYSTEM OF LINEAR VECTOR is related.
In contact point PiInternal unit normal vector can indicate are as follows:
Wherein；eNiThe internal unit normal vector for being i-th of finger in normal direction, gradS (Pi) be contact point Pi gradient,
S (x, y, z)=0 is the shape of target object；
Constraint condition g is seti(y): giIt (y) is the constraint condition for being applied to i-th finger；
gi(y)≤0 for i=1 ..., 2n formula (5)
It is specific:
Wherein；eNi-nAnd foi-nSubscript i value range be i=n+1 ..., 2n, μiIt is static friction coefficient, μsiIt is to turn round
Moment coefficient, E3 are the preset values based on soft finger material and finger structure, and can be in manipulator in crawl target object
It is modified value in the process.N is the quantity of soft finger；B0 and Bi (B1~Bn) is intermediate variable.
A3: the model established based on step A1 and A2 calculates multiple soft fingers to object by Lagrange multiplier
Minimum grip；
It is equivalent by contact force fCThe setting of minimum are as follows:For seeking the minimum value of h (y),
Middle y is vector variable；
H (y)=(W is set#Fr+Ny)T(W#Fr+Ny)；
It introduces a Lagrange multiplier λ and carries out constrained optimization, constraint condition function g (y) and original function h (y) is contacted
To the equation equation equal with variable quantity together, is made into, to find a function minimum of the h (y) under the constraint condition of g (y)=0
Value；Lagrange multiplier λ is indicated are as follows:Obtain the augmentation Lagrange of minimization problem
Function L (y；λ):
L(y；λ)=h (y)+λTG (y) formula (6)
Wherein；L(y；λ) the Augmented Lagrangian Functions to solve minimum value；
The necessary condition for generating the local minimum of h (y) is:
It is obtained from above-mentioned formula:
Wherein；
Obtain y when minimum value h (y)oValue after, be arranged y=yo, the smallest internal force Ny is calculated, and then by formula (3)
Minimal-contact power fc is obtained as minimum grip；
A kind of control method based on soft finger of the invention, can increase to the full extent soft finger and be crawled
Contact area between object reduces grasp force, reduces impact force；Soft finger model is introduced simultaneously, solves chucking power control
System, bring computation complexity when can reduce multiple fingers while accommodating, more particularly to be calculated in the control of real-time power
Moment chucking power is out to carry out finger control.Especially solves the minimum based on the Apery manipulator with multiple soft fingers
Grip algorithmic issue.
The above description is only a preferred embodiment of the present invention, thus it is all according to the configuration described in the scope of the patent application of the present invention,
The equivalent change or modification that feature and principle are done, is included in the scope of the patent application of the present invention.
Claims (8)
1. a kind of control method based on soft finger, which is characterized in that described method includes following steps:
Step 1: robot determines the target object for needing to grab；Specifically: robot task based access control is assigned, and search needs to grab
Target object to be selected；The verification that target object to be selected is carried out after finding target object to be selected, by the mesh to be selected by verification
Mark object is determined as target object；
Robot task based access control is assigned, and search needs the target object to be selected grabbed；Specifically: robot passes through human-computer interaction circle
Face or remote terminal reception task assignment, parse the task of appointment, when task is crawl task, determine the mesh for needing to grab
Object and its essential attribute information and depth attribute information are marked, based on the target to be selected in essential attribute information search working space
Object；Wherein, essential attribute information describes the most basic attribute of target object, and the essential attribute is the attribute for being easy to obtain
Information；
The verification that target object to be selected is carried out after finding target object to be selected will be determined as by the target object to be selected of verification
Target object；Specifically: the verification of target object is carried out based on depth attribute information；Wherein, depth attribute information describes mesh
Mark the detailed attribution information other than primary attribute of object；The depth attribute is further calculated or is started
One or more sensor obtains；
Step 2: obtaining the shape of determining target object；The shape is used in the calculating of minimum grip；
Step 3: robot determines the position of target object, and drives soft finger grip target object；
Step 4: grabbing target object and during subsequent movement, calculate minimum grip in real time, and based on minimum grip come
Carry out the clamping of soft finger；The minimum grip of real-time calculating in step 4, including, step A1: establish linearly connecing for soft finger
Touch type；Step A2: manipulator target object crawl model is established；A3: the model established based on step A1 and A2 passes through glug
Bright day multiplier calculates multiple soft fingers to the minimum grip of object.
2. the control method according to claim 1 based on soft finger, which is characterized in that the step A1: establish soft
Property finger linear contact model, specifically: it is linear between the contact force F and finger deformation quantity d of each soft finger of setting
Relationship；
Wherein: formula (1) uses the central point of each soft finger top surface for the LT local coordinate system of origin；F is soft finger
Contact force between target object, n contact forces of the soft finger n contact point are expressed as f1~fn；E is preset value,
The preset value is related to the material of finger, can rule of thumb or convergence algorithm determines；D be soft finger in the z-direction
Deformation quantity maximum value；Be volume point, indicate the lap of soft finger and target object, the lap because
Contact force cause elastic deformation and formed；K is setting constant；BC is that soft finger is in the z-direction when being contacted with any angle α
Deformation quantity；String when r is soft finger maximum distortion；X1, x2 are range of integration in the x-direction；R is the half of soft finger
Diameter；X, y are x/y plane coordinates.
3. the control method according to claim 2 based on soft finger, which is characterized in that the step A2: establish machine
Tool hand target object grabs model, specifically: A21 establishes manipulator crawl model and A22 establishes soft finger contact model.
4. the control method according to claim 3 based on soft finger, which is characterized in that A21: establishing manipulator crawl
Model, specifically:
It establishes using the mass center of target object as the coordinate system O of the origin of coordinate system；Establish the equation of motion of target object, object
The equation of motion of body is using coordinate system O；
Wherein；
Wherein: fi is i-th of soft finger to be produced in the contact force of i-th contact point namely i-th of finger contact target object
Raw fix, fiy, the resultant force of tri- power of fiz, wherein fix, fiz are the tangential component of fi respectively, and fiy is the normal component of fi；
Fc is the contact force of multiple soft fingers for being indicated with the form of vector to target object,It is the used of target object
Property matrix,It is the position and direction of object,It is Coriolis item,It is gravity item,It is to pull
Hand matrix is able to carry out the conversion of coordinate system depending on the posture of position and object with contact point.
5. the control method according to claim 4 based on soft finger, it is characterised in that: the A22: establishing soft hand
Abutment model, specifically:
fc=W#Fr+ Ny formula (3)
Multiple soft fingers are divided into two parts, W to the contact force of object#FrReferred to as operating force, Ny (i.e. N × y) are productions when grasping
Raw internal force, internal force do not generate displacement；
It is further:
W#Fr=[fo1 T fo2 T … fon T]T
N=[N1 T N2 T … Nn T]T
Wherein:It is the pseudo inverse matrix of W,It is the resultant force acted on target object, uses target
Object coordinates system O, N are the null space matrix of W, and the base of the kernel of W isColumn matrix formed,It is any vector,It is the internal force generated when grasping；foiIt is i-th of finger contact under coordinate system O
Contact force caused by object；NiIt is the N matrix of i-th of finger；Pseudo inverse matrix is the generalized form of inverse matrix；
In contact point PiInternal unit normal vector can indicate are as follows:
Wherein；eNiThe internal unit normal vector for being i-th of finger in normal direction, gradS (Pi) be contact point Pi gradient, S (x,
Y, z)=0 be target object shape；
Constraint condition g is seti(y): giIt (y) is the constraint condition for being applied to i-th finger；
gi(y)≤0 for i=1 ..., 2n formula (5)
It is specific:
Wherein；eNi-nAnd foi-nSubscript i value range be i=n+1 ..., 2n, μiIt is static friction coefficient, μsiIt is torque system
Number, E3 is the preset value based on soft finger material and finger structure, and can be in manipulator in the process for grabbing target object
In be modified value；N is the quantity of soft finger；B0 and Bi (B1~Bn) is intermediate variable.
6. the control method according to claim 5 based on soft finger, which is characterized in that there are three independent for soft finger
Joint, finger tip are hemispherical, one layer of silicone rubber of cladding outside finger tip.
7. the control method according to claim 6 based on soft finger, which is characterized in that grabbed in manipulator target object
In modulus type, establish using the mass center of target object as the coordinate system O of the origin of coordinate system；P is contact point, and P1~Pn is n soft
Property finger contact point, establish using the contact point Pi of soft finger and target object as the local coordinate system Li of origin.
8. the control method according to claim 7 based on soft finger, which is characterized in that the A3: being based on step A1
The model established with A2 calculates multiple soft fingers to the minimum grip of object by Lagrange multiplier, specifically:
It is equivalent by contact force fCThe setting of minimum are as follows:For seeking the minimum value of h (y), wherein y is
Vector variable；
H (y)=(W is set#Fr+Ny)T(W#Fr+Ny)；
It introduces a Lagrange multiplier λ and carries out constrained optimization, constraint condition function g (y) and original function h (y) are related to one
It rises, the equation equation equal with variable quantity is made into, to find a function minimum value of the h (y) under the constraint condition of g (y)=0；
Lagrange multiplier λ is indicated are as follows:Obtain the Augmented Lagrangian Functions of minimization problem
L(y；λ):
L(y；λ)=h (y)+λTG (y) formula (6)
Wherein；L(y；λ) the Augmented Lagrangian Functions to solve minimum value；
The necessary condition for generating the local minimum of h (y) is:
It is obtained from above-mentioned formula:
Obtain y when minimum value h (y)oValue after, be arranged y=yo, the smallest internal force Ny is calculated, and then obtained by formula (3)
Minimal-contact power fc is as minimum grip.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710637965.5A CN107225576B (en) | 2017-07-31 | 2017-07-31 | A kind of control method based on soft finger |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710637965.5A CN107225576B (en) | 2017-07-31 | 2017-07-31 | A kind of control method based on soft finger |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107225576A CN107225576A (en) | 2017-10-03 |
| CN107225576B true CN107225576B (en) | 2019-07-19 |
Family
ID=59957390
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710637965.5A Active CN107225576B (en) | 2017-07-31 | 2017-07-31 | A kind of control method based on soft finger |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107225576B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107696048B (en) * | 2017-10-31 | 2019-12-20 | 华中科技大学 | Rigid-flexible coupling mechanical finger with force position sensing function and sensing method thereof |
| JP6912415B2 (en) | 2018-04-10 | 2021-08-04 | ファナック株式会社 | Hand control device and hand control system |
| CN109227539B (en) * | 2018-09-27 | 2021-12-17 | 北京邮电大学 | Method for minimizing collision force of space flexible mechanical arm |
| CN109635474B (en) * | 2018-12-19 | 2023-04-28 | 杭州电子科技大学 | Method for improving universality of construction of additive cost function in mechanical arm gripping behavior |
| DE102019107975B3 (en) * | 2019-03-28 | 2020-08-13 | Franka Emika Gmbh | Teaching in a holding force for an object in a robotic gripper |
| CN110750867B (en) * | 2019-09-19 | 2021-02-26 | 上海大学 | Overall rigidity and contact force calculation method for soft fingertip structure of hemispherical robot |
| CN110705059B (en) * | 2019-09-19 | 2021-04-27 | 上海大学 | Rigidity identification and contact force calculation method for soft fingertip structure of hemispherical robot |
| CN111015740B (en) * | 2019-12-12 | 2022-05-13 | 中国科学院重庆绿色智能技术研究院 | Touch and slide sensation sensor, flexible finger grabbing system and grabbing method thereof |
| CN114083538B (en) * | 2021-11-30 | 2023-06-20 | 深圳市优必选科技股份有限公司 | Method, device and robot for realizing stable clamping of double mechanical arms |
| CN116587287A (en) * | 2023-06-29 | 2023-08-15 | 廊坊市珍圭谷科技有限公司 | Automatic control method and system for transfer robot |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010005732A (en) * | 2008-06-26 | 2010-01-14 | Toyota Industries Corp | Robot hand mechanism, robot having robot hand mechanism and control method of robot hand mechanism |
| WO2012005834A2 (en) * | 2010-07-08 | 2012-01-12 | Vanderbilt University | Continuum robots and control thereof |
| CN105598995A (en) * | 2016-03-17 | 2016-05-25 | 华东交通大学 | Flexible mechanical arm and control method thereof |
| CN106003131A (en) * | 2016-07-01 | 2016-10-12 | 北京软体机器人科技有限公司 | Double-passage soft finger and soft robot |
| CN106335064A (en) * | 2016-11-29 | 2017-01-18 | 合肥工业大学 | Controller design method for flexible joint robot system |
| CN206123698U (en) * | 2016-10-11 | 2017-04-26 | 佛山科学技术学院 | Gentle and agreeable type no. 3 indicates holder with sense of touch sensing function |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9496080B2 (en) * | 2015-03-24 | 2016-11-15 | The Boeing Company | Flexible electromagnetic clamp |
-
2017
- 2017-07-31 CN CN201710637965.5A patent/CN107225576B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010005732A (en) * | 2008-06-26 | 2010-01-14 | Toyota Industries Corp | Robot hand mechanism, robot having robot hand mechanism and control method of robot hand mechanism |
| WO2012005834A2 (en) * | 2010-07-08 | 2012-01-12 | Vanderbilt University | Continuum robots and control thereof |
| CN105598995A (en) * | 2016-03-17 | 2016-05-25 | 华东交通大学 | Flexible mechanical arm and control method thereof |
| CN106003131A (en) * | 2016-07-01 | 2016-10-12 | 北京软体机器人科技有限公司 | Double-passage soft finger and soft robot |
| CN206123698U (en) * | 2016-10-11 | 2017-04-26 | 佛山科学技术学院 | Gentle and agreeable type no. 3 indicates holder with sense of touch sensing function |
| CN106335064A (en) * | 2016-11-29 | 2017-01-18 | 合肥工业大学 | Controller design method for flexible joint robot system |
Non-Patent Citations (1)
| Title |
|---|
| 机器人柔性抓取试验平台的设计与抓持力跟踪阻抗控制;王学林等;《农业工程学报》;20150131;第31卷(第1期);58-63 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107225576A (en) | 2017-10-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107225576B (en) | A kind of control method based on soft finger | |
| CN109202942A (en) | The simulator of hand control device, hand control method and hand | |
| CN109739358B (en) | Gesture collision detection method and device based on naked eye 3D | |
| Costanzo et al. | Two-fingered in-hand object handling based on force/tactile feedback | |
| CN109794933B (en) | Robot fingertip design method, grabbing planner and grabbing method | |
| JP4001105B2 (en) | Grasping method of arbitrarily shaped object by robot | |
| CN109176590B (en) | Flexible fingertip with pressure and slip feeling perception, device and method | |
| Payeur et al. | Intelligent haptic sensor system for robotic manipulation | |
| CN108818586B (en) | Object gravity center detection method suitable for automatic grabbing by manipulator | |
| CN108453735B (en) | Grabbing method and device based on bionic manipulator of friction nano generator | |
| Ullmann et al. | Intuitive virtual grasping for non haptic environments | |
| CN112428308B (en) | Robot touch action recognition system and recognition method | |
| Su et al. | Robust grasping for an under-actuated anthropomorphic hand under object position uncertainty | |
| CN109693234A (en) | Robot falling prediction method and device, terminal equipment and computer storage medium | |
| AU2016401548B2 (en) | Multi-measurement-mode three-dimensional measurement system and measurement method | |
| CN108227928B (en) | Picking method and device in virtual reality scene | |
| CN115686193A (en) | Virtual model three-dimensional gesture control method and system in augmented reality environment | |
| CN109871116B (en) | Apparatus and method for recognizing gesture | |
| JP2006130580A (en) | Method for gripping arbitrarily-shaped object by robot | |
| JPWO2019069361A1 (en) | Gripping position / posture teaching apparatus, gripping position / posture teaching method, and robot system | |
| JP2013182554A (en) | Holding attitude generation device, holding attitude generation method and holding attitude generation program | |
| CN114882113B (en) | Five-finger mechanical dexterous hand grabbing and transferring method based on shape correspondence of similar objects | |
| CN212084102U (en) | Wearable intelligent glove | |
| CN109901439A (en) | A kind of walking robot control device and method | |
| García-Rodríguez et al. | Dynamic optimal grasping of a circular object with gravity using robotic soft-fingertips |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 150000 Heilongjiang Province Harbin Pingfang District Haping Road Concentrated Area Dalian North Road and Xingkai Road Intersection Applicant after: HIT ROBOT GROUP Co.,Ltd. Address before: 150000 Heilongjiang Province Harbin Pingfang District Haping Road Concentrated Area Dalian North Road and Xingkai Road Intersection Applicant before: HIT ROBOT GROUP Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PP01 | Preservation of patent right | ||
| PP01 | Preservation of patent right |
Effective date of registration: 20240626 Granted publication date: 20190719 |