Information Technology Reference
In-Depth Information
,
, 1
,
,
,
,
, 2
, 3
(17)
where
,
is the
j-th
dimensional velocity of bat
i
at iteration
t
and is modified by a
random value,
, from a uniform distribution.
is the new location of bat by three
designed strategies. As a result, a new found location consist of different elements
from position, location and random for the bat to deviate the search direction of bats.
The divers search benefits the bats to exploit new search region to avoid falling into a
local optimal solution. Then, the cosine similarity is utilized to analyze the similarity
of the new location and original location of a bat as Eq. (18).
∑
∑
_
,
(18)
∑
_
is the similarity of a new location and a current location. If the similarity of
bat is greater than the similarity threshold (
), the bat is still search the location
around their own location after the search direction of bat suffers violent disturb by
designed strategies. This bat will discover a new location through tracking for the
current best bat to fast approach to the current best bat as shown in Eq. (19).
is
the similarity threshold.
,
1
,
(19)
Step 6. Update Bat Information
This bat will update its properties to afresh adjust status, including loudness (
A
i
) and
emission pulse rate (
r
i
) according to Eq. (20). This new current best bat with better
status will lead all bats toward new search direction to explore new area. The loudness
and emission pulse rate of bats will be updated when (1) the current best bat is up-
dated in divers search and (2) the random number (
r
A
) between 0 and 1 is smaller than
the loudness (
A
i
) of bat
i
.
,
1
(20)
Step 7. Termination
All bats carry out Step 2-6 to indicate that iterations have been finished. All bats will
be back to Step 2 to start next iteration repeatedly until the condition of termination is
satisfied. When the termination is satisfied, the final current best bat is regarded as the
global best bat and its position is the derived global optimal solution of the optimiza-
tion problem in this evolution.
Search WWH ::
Custom Search