Scribd
Upload
35 views75 pages

Lecture 6 - Manipulators Velocities

The document discusses the differential kinematics of manipulators, focusing on the relationship between joint velocities and end-effector velocities. It covers concepts such as velocity transformation, Jacobians, and the propagation of velocities through a chain of robotic links. Additionally, it addresses the challenges of computing velocities for multi-joint robotic arms and introduces methods for resolved-rate motion control in 2D and 3D environments.

Uploaded by

Waleed Nasir
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
35 views75 pages

Lecture 6 - Manipulators Velocities

The document discusses the differential kinematics of manipulators, focusing on the relationship between joint velocities and end-effector velocities. It covers concepts such as velocity transformation, Jacobians, and the propagation of velocities through a chain of robotic links. Additionally, it addresses the challenges of computing velocities for multi-joint robotic arms and introduces methods for resolved-rate motion control in 2D and 3D environments.

Uploaded by

Waleed Nasir
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 75

Manipulator Velocities

Dr. Sara Ali


Introduction
• Previously we discussed the relationship between joint coordinates
and end-effector pose – the manipulator kinematics.
• Now we investigate the relationship between the rate of change of
these quantities – between joint velocity and velocity of the end-
effector.
• This is called the velocity or differential kinematics of the manipulator.
Introduction
• Forward Kinematics

• Inverse Kinematics

• Velocity Transform
Differentiation of vector
• We see that we are calculating the derivative of Q relative to frame
{B}. For example,
• if Q is not changing in time relative to {B}, then the velocity calculated is zero
—even if there
• is some other frame in which Q is varying. Thus, it is important to indicate the
frame in which the vector is differentiated.
Two important frames
• Velocity vector is associated with a point in space
• BUT the numerical values describing the velocity of that point depend on two
frames:
• one with respect to which the differentiation was done
• and one in which the resulting velocity vector is expressed.
Defining velocity w.r.t origin
• Rather than considering a general point's velocity relative to an arbitrary frame,
consider the velocity of the origin of a frame relative to some understood
universe reference frame

• the notation to refer to the velocity of the origin of frame {C}; then the
velocity of the origin of frame {C} expressed in terms of frame {A} (though
differentiation was done relative to {U}).
• Linear velocity describe an attribute
of a point
• Angular velocity describes an
attribute of a body

• The direction of A indicates the instantaneous axis of rotation of {B} relative to {A}, and the magnitude
indicates the speed of rotation.
Relative orientation of B
with respect to A is
constant

Only that case in which relative orientation of (B) and (A} remains constant.
Two frames with coincident origins and with
zero linear relative velocity; their origins will
remain coincident for all time.
Reading Asg
• Chapter 5, Figure 5.5, reading asg, page 140
Simultaneous linear and rotational velocity

• We can very simply expand the previous equation to the case where origins are
not coincident by adding on the linear velocity of the origin to derive the general
formula for the velocity of a vector fixed in frame {B} as seen from frame {A}:

Final result for the derivative of a vector in a moving frame as seen from a stationary frame.
Motion of Links of Robot

A manipulator is a chain of bodies, each one capable of motion relative to its neighbors. Because of this
structure, we can compute the velocity of each link in order, starting from the base.
Velocity Propagation from link
to link
Velocity Propagation from link
to link
• Calculating linear and angular velocities of links of robot
• Manipulator is a chain of rigid bodies moving relative to each other,
therefore,
• Computing velocity of each link in order starting from base.

• So, ?
Velocity Propagation from link
to link
• Therefore, the angular velocity of link i + 1 is the same as that of link i
plus a new component caused by rotational velocity at joint i + 1. This
can be written in terms of frame {i} as
The angular velocity of link I + 1 is the
same as that of link i plus a new component
caused by rotational velocity at joint i + 1.

The rotation matrix rotates the axis of rotation of


joint i + 1 into its description in frame {i so that the
two components of angular velocity can be added.

P.S: Rotational velocities can be added when both


cv vectors are written with respect to the same
frame
Velocity Propagation
• For prismatic joints??

Applying these equations successively from link to link, we can compute the rotational and linear
velocities of the last link.
Jacobian
Jacobian
Jacobian
Asg#4
Asg#4
Velocity of 2-joint Planar Robot Arm
• Relationship between velocity of each individual joint and velocity of
the end effector
Velocity of 2-joint Planar Robot Arm

Joint angles are now a function of time

We can now calculate the derivative using chain


Note that rule

Chain rule
Velocity of 2-joint Planar Robot Arm

Writing above expression in Matrix Form


Inverting Jacobian

Problem?
Singular Pose
Velocity of 3-joint Planar Robot Arm
• Consider Translational Part

Taking Derivative and arranging in matrix form, we get

Dimensions?
Velocity of 3-joint Planar Robot Arm
• Consider Rotational Part
Resolved-rate motion control in 2D
• We are interested in robot end-effector moving in cartesian space
along a straight line. This can be achieved through resolved-rate
motion control.
Resolved-rate motion control in 2D
• Specify the Cartesian Velocity
1. Choose the cartesian velocity v
2. Find the joint velocities using

3. Move the joints at the calculated


speed
4. At next time instant, since q has
changed recompute the jacobian and
update it.
5. Repeat the above process
Resolved-rate motion control in 2D
• The motion controllers work at discrete time steps
• Consider discrete time steps k= 0 ,1,2, 3……..
• We update computation after every Δt seconds
• At each time step
• compute joint velocity
• compute next joint configuration
• Move joint to next configuration
Velocity of 3-joint Planar Robot Arm
Velocity of 3-joint Planar Robot Arm
Forwards Kinematics

Rotational Part Translational Part


Velocity of 3-joint Planar Robot Arm

Spatial Velocity Jacobian Joint Velocity


Velocity of 3-joint Planar Robot Arm
Velocity Kinematics in 3D
• Translational and rotational motion in 3D
• Rate of change of position

• Rate of change of Orientation

• Magnitude is the rate of rotation


• Direction is the axis about which it is rotating
Velocity Kinematics in 3D
• Instantaneous velocity in 3D
is described by 6
parameters

• Also known as spatial


velocity or twist
Velocity of 6-Joint Robot Arm (Translation)

• Earlier for 2-Joint Robot


Velocity of 6-Joint Robot Arm (Translation)

• Earlier for 3-Joint Robot


Velocity of 6-Joint Robot Arm (Translation)

• What about a 6-joint robot such as puma560?


• Very difficult
• Alternative?
• Decouple joints i.e. each joint movement (one at a time) results in different
end-effector movement
• Use approximation
Velocity of 6-Joint Robot Arm (Translation)

• Approximation of Derivative

• In our case
Velocity of 6-Joint Robot Arm (Translation)

After slightly displacing joint 1


Velocity of 6-Joint Robot Arm (Translation)
Computing approximate partial derivative

*We can do the same for remaining joints 2,3,4,5,6


Velocity of 6-Joint Robot Arm (Translation)
Computing approximate partial derivative

*We can do the same for remaining joints 2,3,4,5,6


Skew-Symmetric Matrix Recap

In 2 dimensions the skew-symmetric matrix is


Velocity of 6-Joint Robot Arm (Rotation)

Rearranging the above expression


to isolate skew-symmetric matrix
Velocity of 6-Joint Robot Arm (Rotation)

• Lets check if it is really a skew symmetric matrix and what advantage


can we take from it
Velocity of 6-Joint Robot Arm

• Combining translation and rotation


Velocity of 6-Joint Robot Arm
Inverting Jacobian Matrix

• Jacobian can be inverted unless


singular
• It helps to find the set of joint
velocities necessary to achieve a
desired spatial velocity
• In puma560, there is a singularity
when q5=0
Inverting Jacobian Matrix

*Therefore check determinant of Jacobian


first and then calculate inverse Jacobian
Singular Jacobian

• A Singular Jacobian
• Note columns 4 and 6 are identical, therefore motion of joint 4 or joint 6
results in the same motion (rotation about z-axis)

• Mathematically?
Transforming Spatial Velocity between
Frames

*Jacobian maps velocity from one frame to another and it is a


function of relative orientations of the frames
Jacobian Shape and Number of
Joints
• Number of rows?
• Number of columns?
Jacobian Shape and Number of
Joints
• Underactuated System with 4 joints

Shape and Problem?


Jacobian Shape and Number of
Joints
• Overactuated robot with 100 joints

Shape and Problem?


Jacobian Shape and Number of
Joints
• A pseudoinverse is a matrix inverse-like object that may be defined for a
complex matrix, even if it is not necessarily square. For any given complex
matrix, it is possible to define many possible pseudoinverses.

• Inversion problem is solved through pseudo inversion

Shape and Problem?


100x1=(Nx6) (6x1). Where N would be 100. so J(q)+ will be 100x1. can prove it with second formula too of JJJ
Jacobian Shape and Number of
Joints
• Null space motion (motion of joints that does not effect the end-point
pose)
• Null space

*Set of vectors that satisfy the below condition forms the null space
Null Space Motion

Example
NN+qns will from the set of redundant joints that will not effect the pose of end effectoe but will only change the shape.
(100x1)=(100x6)(6x1) + (100x94)(94x100)(100x1)

You might also like