3/23/2023

MEHRAN UNIVERSITY
OF ENGINEERING & TECHNOLOGY, JAMSHORO
MECHANICAL ENGINEERING, DEPARTMENT
B.E MECHATRONIC ENGINEERING PROGRAM

ROBOTICS

Path & Trajectory

Planning
1

Find joint velocities

 Control a manipulator as shown in figure:
 FIND THETA-1 and
 THETA-2

1
 3/23/2023

Find joint velocities

 Control a manipulator as shown in figure:

Find joint velocities

 Control a manipulator as shown in figure:
 Note: 3 is prismatic joint.

2
 3/23/2023

Path and Trajectory Planning

 In previous lectures, we control the manipulator

in open space. There is no any obstacles.
 Now, here we assume obstacles in the systems.
 We will assume that the initial and final
configurations of the robot are specified and that
the problem is to find a collision free path
connecting these configurations.
 Path and Trajectory planning is a procedure to
find collision free paths for the robot.

Path Planning
 Path planning for industrial robot manipulator is
a very important task that has to be done correct
to avoid damage on personnel, other equipment
in the work area or the robot manipulator itself.
 Path Planning identifies the points from which
the end effector will pass.
 There might be tight spaces, humans working in
same environment or other challenges that the
manipulator has to be aware of in order to find a
collision free path.
 It requires a lot of computational power when
solving a path planning problem because there
can be infinitely many paths between the starting
position and goal position of the robot.

3
 3/23/2023

Trajectory Planning
 The trajectory planner computes a function q(t)
that completely specifies the desired motion of
the robot as it traverses the path.
 Trajectory Planning identifies the velocity
components of the end effector motion along the
path.
 Path planning for industrial robot manipulator is
a very important task that has to be done correct
to avoid damage on personnel, other equipment
in the work area or the robot manipulator itself.

Path and Trajectory Planning

 Here if end effector will move from (X1,Y1) to
(X2,Y2), the Path will be a straight line.
 A line equation will be used to generate a path
and this equation is known as parametric
equation, as it depends on time(t).

 Where x is distance, y is distance, and v is

total/resultant velocity

4
 3/23/2023

Path and Trajectory Planning

Path and Trajectory Planning

 As we know that if joint velocities are given then how linear velocities of end
effector can be found. 𝑥
𝑦
𝑉 𝑧
𝜁 =
𝜔 𝜔
𝜔
𝐽 𝐿𝑖𝑛𝑒𝑎𝑟 𝑉𝑒𝑙𝑜𝑐𝑖𝑡𝑖𝑒𝑠 𝜔
𝐉
𝐽 𝐴𝑛𝑔𝑢𝑙𝑎𝑟 𝑉𝑒𝑙𝑜𝑐𝑖𝑡𝑖𝑒𝑠

 Similarly, if end effector velocities are given how we find joint velocities.

10

5
 3/23/2023

Path and Trajectory Planning

 In the path planning a complete
specification of the location of every
point on the robot is referred to as a
configuration, and the set of all
possible configurations is referred to
as the configuration space.
 For our purposes, the vector of joint
variables q provides a convenient
representation of a configuration. The
configuration space is denoted by Q.

11

THE CONFIGURATION SPACE

 In one link revolute arm Q=S1, where S1 represents the unit circle. a
configuration by q = 1.
 In two links revolute arm Q=S1 x S1 = T2, where T2 represents the unit
Torus. a configuration by q = (1, 2).
 A Cartesian arm, Q =R3, and configuration by q = (d1, d2, d3) = (x, y, z).
 A collision occurs when the robot contacts an obstacle in the workspace.
 To describe collisions, some additional notation are introduced.
• A is the robot
• A(q) the subset of the workspace that is occupied by the robot at
configuration q.
• Oi is the obstacles in the workspace, and
• W is the workspace (i.e., the Cartesian space in which the robot
moves).

12

6
 3/23/2023

THE CONFIGURATION SPACE

 Rigid body is showing here as end-effector.

 Obstacle is rectangular shape.
 Motion of end-effector is shown by dotted lines.

13

Trajectories for Point to Point Motion

 Here find a trajectory that connects an initial to a final configuration while satisfying
other specified constraints at the endpoints (e.g., velocity and acceleration).
 At time t0 the joint variable satisfies q(t0)=q0, q’(t0)=q’0 and q”(t0)=0
 At time tf the joint variable satisfies q(tf)=qf, q’(tf)=q’f and q”(tf)=f
 Figure shows a suitable trajectory for this motion.

14

7
 3/23/2023

Trajectories for Point to Point Motion

 Cubic Polynomial Trajectories:
 To generate a trajectory between two configurations, the start and end velocities for the trajectory
must be specified.
 One way to generate a smooth curve is by a polynomial function of t.
 A cubic trajectory of the form:
 Then the desired velocity is automatically given as:
 Combining equations these with the four constraints yields four equations in four unknowns;

 Write it in matrix form:

15

Trajectories for Point to Point Motion

 Cubic Polynomial Trajectories:
 Step # 1 Find Constants (a0, a1, a2 and a3) between 2 points (A and B).
 Note: These constants are valid between these 2 points (A and B).
 Step # 2 Develop equation by putting values of (a0, a1, a2 and a3).
 Step # 3 Drive equation and find q(t) and q’(t).
 Step # 4 Make table for time (0 to max) where t, q(t), and q’(t) are columns of table.
 Draw / plot Displacement and Velocity diagrams

16

8
 3/23/2023

Trajectories for Point to Point Motion

 Cubic Polynomial Trajectories:
 Problem: Generate Cubic Polynomial Trajectory where the initial and final velocities
are zero. Take t0 = 0 and tf = 1 sec

A required polynomial trajectory function is:

17

Trajectories for Point to Point Motion

 Cubic Polynomial Trajectories:

 Problem: Generate Cubic Polynomial Trajectory where the initial and final velocities
are zero. Take t0 = 0 and tf = 1 sec
 A- Assume qo=10 degrees and qf=-20 degrees
 B- Assume qo=14 degrees and qf=26 degrees

18

9
 3/23/2023

Trajectories for Point to Point Motion

 Multiple Cubics Polynomial Trajectories:

 A sequence of moves can be planned using the above
formula by using the end conditions qf , vf of the i-th
move as initial conditions for the i + 1-st move.
 In other words, the final velocity of previous move is the
initial velocity of next move
 Problem: Figure shows a 6-second move, computed in
three parts, where the trajectory begins:
 1- At 10o and is required to reach 40o at 2-seconds,
 2- At 40o and is required to reach 30o at 4 seconds, and
 3- At 30o and is required to reach 90o at 6-seconds, with
zero velocity at 0,2,4, and 6 seconds.

19

Trajectories for Point to Point Motion

 Multiple Cubics Polynomial Trajectories:

 A sequence of moves can be planned using the above
formula by using the end conditions qf , vf of the i-th
move as initial conditions for the i + 1-st move.
 In other words, the final velocity of previous move is the
initial velocity of next move
 Problem: Figure shows a 6-second move, computed in
three parts, where the trajectory begins:
 1- At 10o and is required to reach 40o at 2-seconds,
 2- At 40o and is required to reach 30o at 4 seconds, and
 3- At 30o and is required to reach 90o at 6-seconds, with
zero velocity at 0,2,4, and 6 seconds.

20

10
 3/23/2023

Trajectories for sequence of Motion

 Quintic Polynomial Trajectories:

 The planning trajectories using multiple cubic trajectories leads to
continuous positions and velocities at the blend times but
discontinuities in the acceleration.
 The derivative of acceleration is called the Jerk. A discontinuity in
acceleration leads to an impulsive Jerk, which may excite vibrational
modes in the manipulator and reduce tracking accuracy.
 For this reason, one may wish to specify constraints on the
acceleration as well as on the position and velocity. In this case, we
have six constraints. (Position, velocity, acceleration at initial and
final).
 Therefore a fifth order polynomial is required.

 Take 1st derivate for velocity and 2nd derivative for acceleration we get:
21

Trajectories for sequence of Motion

 Quintic Polynomial Trajectories:

 Take 1st derivate for velocity and 2nd derivative for acceleration
we get:

22

11
 3/23/2023

Trajectories for sequence of Motion

 Quintic Polynomial Trajectories:

 Take 1st derivate for velocity and 2nd derivative for acceleration
we get:
 In matrix form:

23

Trajectories for sequence of Motion

 Quintic Polynomial Trajectories:

 Problem: Generate Quintic Polynomial Trajectory where the initial and
final velocities and accelerations are zero. Take t0 = 0, t1 = 2, t2 = 4, and
t3 = 6 sec.

24

12
 3/23/2023

Trajectories for sequence of Motion

 Problem-A: Generate a trajectory of a 6-axis robot, where the first joint move from an initial angle
of 50 to a final angle of 80 in 3 seconds. Calculate the coefficients for a third-order polynomial
joint-space trajectory. Determine the joint angles, velocities, and accelerations at 1, 2, and 3
seconds.
 Problem-B: Generate a trajectory of a 6-axis robot, where the second joint of a 6-axis robot is to
move from initial angle of 20 to an middle angle of 80 in 5 seconds and continue to its
destination of 25 in another 5 seconds. Calculate the coefficients for third-order polynomials in
joint-space. Plot the joint angles, velocities, and accelerations. Assume the joint stops at
intermediate points.
 Problem-C: Generate a trajectory, a fifth-order polynomial is to be used to control the motions of
the joints of a robot in joint-space. Find the coefficients of a fifth-order polynomial that will allow a
joint to go from an initial angle of 0 to a final joint angle of 75 in 3 seconds, while the initial and
final velocities are zero and initial acceleration and final decelerations are 10 / sec2 .

25

BOOKS

1- Introduction to Robotics: Analysis, Control, Applications

By Saeed Benjamin Niku
Latest edition

2- Introduction to Robotics Mechanics and Control

By John J. Craig
3rd Edition

3- Theory of Applied Robotics

Kinematics, Dynamics and Control,
By Reza N. Jazar
2nd edition

4- Robotics, Vision and Control,

Fundamental Algorithms in Matlab
By Peter Corke
2nd Edition
26

26

13