Particle Swarm Optimization - PowerPoint PPT Presentation

1 / 10
About This Presentation
Title:

Particle Swarm Optimization

Description:

Inspired by the social behavioural patterns of organisms along with some physics ... than trying to impose it, that emulates nature rather than trying to control it. ... – PowerPoint PPT presentation

Number of Views:82
Avg rating:3.0/5.0
Slides: 11
Provided by: Namrata1
Category:

less

Transcript and Presenter's Notes

Title: Particle Swarm Optimization


1
Particle Swarm Optimization
  • Winter 2005

2
Introduction
  • Developed by Dr. Russ Eberhart (Electrical
    Engineer, Purdue School of Engineering and
    Technology) and Dr. James Kennedy (social
    psychologist, Bureau of Labor Stats) in 1995
  • Easy to Implement
  • Inspired by the social behavioural patterns of
    organisms along with some physics based
    interactions

3
Heppners Birds
  • Birds attracted to a roosting area
  • Entire flock eventually lands on the roost
  • Finding the roost is similar to finding a
    solution in a field of possible solutions

4
PSO
  • Learn from the successes of neighbours
  • Try to achieve a balance between exploration
    (global search) and exploitation (local search)

5
PSO Basic Algorithm
  • 1. Initialize the population (assign velocities
    and locations)
  • 2. Evaluate the fitness of individual particles
  • 3. Keep track of location where the individual
    had its highest fitness
  • 4. Modify velocities based on previous best and
    global best positions.
  • 5. Update the particles position
  • 6. Terminate if some condition is met
  • 7. Go to Step 2

6
Detailed Algorithm
  • 1. Initialize the population. Each individual is
    defined by its locations and its velocity p
    locations vector
  • v velocities vector

7
Detailed Algorithm Cont.
  • 2. Evaluate the fitness of individual particles
  • 3. Keep track of location where the individual
    had its highest fitness

8
Detailed Algorithm Cont.
  • 4. Modify velocities based on pBest and gBest
    positions. and are constants
  • 5. Update the particles position
  • 6. Terminate if some condition is met
  • 7. Go to step 2
  • Note Random Number is also added to the
    particles velocity and its position

9
Parameters
  • Alpha is the inertia weight
  • V_max

10
Conclusion
  • PSO tries to incorporate both evolutionary and
    swarm aspects
  • "PSO belongs ideologically to the philosophical
    school that allows wisdom to emerge rather than
    trying to impose it, that emulates nature rather
    than trying to control it." - Eberhart and
    Kennedy
Write a Comment
User Comments (0)
About PowerShow.com