Planning Techniques for Robotics

Mahmoud Abdelaal

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 1

 Lecture # 5

Particle Swarm Optimization I

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 2

 Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 3

Introduction to PSO
Introduction to the PSO: Origins
• Inspired from the nature social behavior and dynamic movements with
communications of insects, birds and fish
• Proposed by James Kennedy & Russell Eberhart (1995)
• Swarm is better than personal

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 4

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 5

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 6

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 7

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 8

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 9

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 10

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 11

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 12

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 13

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 14

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 15

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 16

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 17

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 18

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 19

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 20

Introduction to PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 21

Introduction to PSO
The basic idea
• Each particle is searching for the optimum
• Each particle is moving and hence has a velocity.
• Each particle remembers the position it was in where it had its best result so
far (its personal best)
• But this would not be much good on its own; particles need help in figuring
out where to search.
• The particles in the swarm co-operate. They exchange information about what
they’ve discovered in the places they have visited

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 22

Introduction to PSO
The Concept of PSO
• Uses a number of agents (particles) that constitute a swarm moving around in
the search space looking for the best solution
• Each particle in search space adjusts its “flying” according to its own flying
experience as well as the flying experience of other particles

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 23

Introduction to PSO
The Concept of PSO
The idea is similar to bird flocks searching for food.
• Bird = a particle,
• Food = a solution
• pbest = the best solution (fitness) a particle has achieved so far.
• gbest = the global best solution of all particles within the swarm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 24

Introduction to PSO
The Concept of PSO
• Each particle modifies its position according to:
• its current position
• its current velocity
• the distance between its current position and pbest
• the distance between its current position and gbest

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 25

Introduction to PSO
The Concept of PSO
• Collection of flying particles (swarm) - Changing solutions
• Search area - Possible solutions
• Movement towards a promising area to get the global optimum
• Each particle keeps track:
• its best solution, personal best, pbest
• the best value of any particle, global best, gbest

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 26

Introduction to PSO
PSO Search Scheme
• The basic concept of PSO lies in accelerating each particle toward its pbest
and the gbest locations, with a random weighted acceleration at each time.
• PSO uses a number of agents (particles), that constitute a swarm flying in the
search space looking for the best solution.
• Each particle is treated as a point (candidate solution) in a space which
adjusts its “flying” according to its own flying experience as well as the
flying experience of other particles.
• The neighborhood concept in PSO is not the same as the one used in other
meta-heuristics search, since in PSO each particle’s neighborhood never
changes (is fixed)

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 27

 Mathematical Model of PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 28

Mathematical Model of PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 29

Mathematical Model of PSO

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 30

Mathematical Model of PSO
Each particle tries to modify its position X using the following formula:

X (t+1) = X(t) + V(t+1) (1)

V(t+1) = wV(t) + c1 ×rand ( ) × ( Xpbest - X(t)) + c2 ×rand ( ) × ( Xgbest - X(t)) (2)

V(t) velocity of the particle at time t

X(t) Particle position at time t
w Inertia weight
c1 , c2 learning factor or accelerating factor (Constants)
rand uniformly distributed random number between 0
and 1
Xpbest particle’s best position
Xgbest global best position

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 31

Mathematical Model of PSO
PSO Algorithm - Parameters

• Number of particles usually between 10 and 50

• C1 is the importance of personal best value
• C2 is the importance of neighborhood best value
• Usually C1 + C2 = 4 (empirically chosen value)
• If velocity is too low → algorithm too slow
• If velocity is too high → algorithm too unstable

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 32

Mathematical Model of PSO
• Particle’s velocity:

1. Inertia • Makes the particle move in the same

direction and with the same velocity

2. Personal • Improves the individual

• Makes the particle return to a previous
Influence
position, better than the current

3. Social • Makes the particle follow the best

Influence neighbors direction

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 33

 PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 34

PSO Algorithm
PSO Algorithm
1) Create a ‘population’ of agents (particles) uniformly distributed over X
2) Evaluate each particle’s position according to the objective function
3) If a particle’s current position is better than its previous best position,
update it
4) Determine the best particle (according to the particle’s previous best
positions)
5) Update particles’ velocities:
6) Move particles to their new positions:
7) Go to step 2 until stopping criteria are satisfied

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 35

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 36

PSO Algorithm
The PSO algorithm pseudocode as following:
Input: Randomly initialized position and velocity of Particles:
Xi (0) andVi (0)
Output: Position of the approximate global minimum X*

1: while terminating condition is not reached do

2: for i = 1 to number of particles do
3: Calculate the fitness function f
4: Update personal best and global best of each particle
5: Update velocity of the particle using Equation 2
6: Update the position of the particle using equation 1
7: end for
8: end while

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 37

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 38

PSO Algorithm
• Simulation 1

max
y

min
fitness

x
search space

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 39

PSO Algorithm
• Simulation 2

max
y

min
fitness

x
search space

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 40

PSO Algorithm
• Simulation 3

max
y

min
fitness

x
search space

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 41

PSO Algorithm
• Simulation 4

max
y

min
fitness

x
search space

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 42

PSO Algorithm
• Simulation 5

max
y

min
fitness

x
search space

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 43

PSO Algorithm
• Simulation 6

max
y

min
fitness

x
search space

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 44

PSO Algorithm
• Simulation 7

max
y

min
fitness

x
search space

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 45

PSO Algorithm
• Simulation 8

max
y

min
fitness

x
search space

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 46

PSO Algorithm
• Search result

Iteration Swarm best

0 416.245599
5 515.748796
10 759.404006
15 793.732019
20 834.813763
100 837.911535
5000 837.965771
Global 837.9658

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 47

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 48

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 49

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 50

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 51

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 52

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 53

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 54

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 55

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 56

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 57

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 58

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 59

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 60

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 61

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 62

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 63

PSO Algorithm

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 64

PSO Algorithm
Algorithm Characteristics
Advantages
• Insensitive to scaling of design variables
• Simple implementation
• Easily parallelized for concurrent processing
• Derivative free
• Very few algorithm parameters
• Very efficient global search algorithm
Disadvantages
• Tendency to a fast and premature convergence in mid optimum points
• Slow convergence in refined search stage (weak local search ability)

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 65

Thank you for your attention !

Good Luck !

Mahmoud Abdelaal Planning Techniques for Robotics_@_BFCAI 66