Page 13 - Algorithm Collections for Digital Signal Processing Applications using MATLAB

Chapter 1

ARTIFICIAL INTELLIGENCE
Algorithm Collections

1. PARTICLE SWARM ALGORITHM

Consider the two swarms flying in the sky, trying to reach the particular
destination. Swarms based on their individual experience choose the proper
path to reach the particular destination. Apart from their individual
decisions, decisions about the optimal path are taken based on their
neighbor’s decision and hence they are able to reach their destination faster.
The mathematical model for the above mentioned behavior of the swarm is
being used in the optimization technique as the Particle Swarm Optimization
Algorithm (PSO).
For example, let us consider the two variables ‘x’ and ‘y’ as the two
swarms. They are flying in the sky to reach the particular destination (i.e.)
2
they continuously change their values to minimize the function (x-10) +(y-
2
5) . Final value for ‘x’ and ‘y’ are 10.1165 and 5 respectively after 100
iterations.
The Figure 1-1 gives the closed look of how the values of x and y are
changing along with the function value to be minimized. The minimization
function value reached almost zero within 35 iterations. Figure 1-2 shows
the zoomed version to show how the position of x and y are varying until
they reach the steady state.

1