Вы находитесь на странице: 1из 2

Particle Swarm Optimization

History

Particle Swarm Optimization (PSO) was designed in 1995 by the Kennedy and Eberhart. They designed it
for the simulation of social behavior of birds and fishes by using computer graphics and social
psychology research. They take forward the work of Reeves who introduces the particle system which
uses object models that are dynamic and cannot be easily represented by polygons or surfaces.
Examples of such objects are clouds, fire, water and smoke. In the field of computer graphics Reevas
was the first who introduces the first particle system in 1983.

In the particle system which is introduces by Reevas particles are independent of each other and their
movement is governed by some specific set of rules. Later on in 1987, Reynolds uses this particle system
to simulate the behavior of flocks of birds. After that in 1990 Heppner and Grenandor copied the system
of Reynolds but with the addition of roost in his system which is helpful in attracting the birds. Both of
these models followed the set of rules and later these were used in the original algorithm (Kiranyaz,
Ince, & Gabbouj, 2014).

Introduction

Particle swarm Optimization (PSO) is a population-based algorithm. It was invented by Russel Eberthat
and James Kennedy in 1995. It is basically designed for the schooling and flocking pattern of birds and
fishes but later it was found that its algorithm is best described with optimization process. Its algorithm
is similar to the generic algorithm. A collection of individuals called particles move in it through a region.
In the field of computer science it is thinks of as a computational method that founds the solution of
problems by improving candidate solution with respect to measure of given quality.

PSO is not only used for optimization but also used for representing sociology of humans and artificial
agents which are based on social psychology. In order to balance exploration and exploitation PSO is
used to combine local search method with global search method. In PSO particles move around in a
multidimensional search space. During flight movement every particle adjust their desired position
according to its own experience and through the experience of neighboring particle they can adjust their
best position by itself and its neighbor.

How it Works

PSO is initialized with a group of random particles and searches for optima by updating generations. In
PSO particles moves through the solution space and are evaluated according to some fitness criterion
after each time step. In every attempt, each particle is updated by following two “best” values. The first
one is the best solution (fitness) it has achieved so far (the fitness value is also stored). This value is
known as pbest. Another “best” value that is tricked by the particle swarm optimization is the best value
obtained by any particle in the population. This second best value is known as gbest.

Particle Swarm Optimization has quality factors of pbest and gbest. Particle Swarm Optimization has a
diverse response between pbest and gbest. It is an artificial intelligence (AI) technique that is used to
find approximate solution to extreme difficult. It has a simple algorithm which is easy to implement and
few parameter are used to maintain the velocity. Its parameters are population size, velocity etc.
Applications

The applications of Particle Swarm Optimization are as follows.

 Image and video analysis.


 Design and Restructuring of electricity networks and load dispatching.
 Control Applications.
 Electronics and Electromagnetics applications.
 Antenna design.
 Power generation and Power systems.
 Design Applications.
 Design and optimization of communication networks.
 Biological, Medical and Pharmaceutical.
 Clustering, Classification and data mining.
 Fuzzy and neuron-fuzzy systems and control.
 Signal Processing.
 Neural Networks.
 Combinatorial optimization problems.
 Robotics.
 Prediction and forecasting.
 Modeling
 Detection and diagnosis of faults and recovery for them.
 Sensors Networks.
 Applications in metallurgy.
 Music generation and games.
 Security and military applications.
 Finance and economics.

Вам также может понравиться