The
performance of firefly algorithm is affected by the values of α, γ and β
parameters. The value of α varies from 0 to 1. Its value is generated randomly
between 0 and 1 which adds randomness to the algorithm. The value of γ can vary
from [0
. If γ
is considered 0 that means the attractiveness is constant between fireflies and
there is no movement of fireflies which may give single optimum local maxima.
If γ=∞ then there is no attraction i.e. sight of other fireflies is short-sighted
with respect to other fireflies which leads each firefly to roam in a completely
random way [19-21]. To avoid this, value of γ is varied between 0.01 and 100 in
an iterative manner which finally results γ=1 while β0 is taken as
1. In this paper, α is varied as number of iterations increases as in PSO [32-33].