In real-coded genetic algorithms (GAs), some crossover operators do not wor
k well on functions which have their optimum at the corner of the search sp
ace. To cope with this problem, we have proposed a boundary extension metho
d which allows individuals to be located within a limited space beyond the
boundary of the search space. In this paper, we give an analysis of the bou
ndary extension methods from the viewpoint of sampling bias and perform a c
omparative study on the effect of applying two boundary extension methods,
namely the boundary extension by mirroring (BEM) and the boundary extension
with extended selection (BES). We were able to confirm that to use samplin
g methods which have smaller sampling bias had good performance on both fun
ctions which have their optimum at or near the boundaries of the search spa
ce, and functions which have their optimum at the center of the search spac
e. The BES/SD/A (BES by shortest distance selection with aging) had good pe
rformance on functions which have their optimum at or near the boundaries o
f the search space. We also confirmed that applying the BES/SD/A did not ca
use any performance degradation on functions which have their optimum at th
e center of the search space. (C) 2001 Elsevier Science Inc. All rights res
erved.