Geology Reference
In-Depth Information
Figure 4. Flow chart of NS2-IRR GA (Cha 2008)
SP2-IRR GA
P
t
+1
the crowding distances are assigned to the
best remedy rank
F
i
using Equation (5). By add-
ing
F
i
to
P
t
+1
and sorting in descending order by
using Equation (6), the first N number of indi-
viduals is selected. For the fourth step,
Q
t
+1
is
generated by crossover and mutation operator.
This iteration will continue until satisfying the
GA criteria or reaching the predefined maximum
generation. Although the performance of the NS2-
IRR GA is good, it is often difficult to guarantee
diverse optimal solution sets. However, a solution
can be found in the modified strength Pareto
evolutionary algorithm.
Strength Pareto Evolutionary
Algorithm (SPEA)
Zitzler and Thiele (1999) proposed the strength
Pareto evolutionary algorithm (SPEA) with com-
bination of several features of multi-objective
evolutionary algorithms in a unique manner.
The SPEA has some similarities in its process to
other evolutionary algorithms (Zitzler and Thiele
1999) in that it
Search WWH ::
Custom Search