1- We ¯rst randomly choose two particles out of the population which
excludes the particle whose velocity is updated.
2- We compare the ¯tness values of these two particles' pbest's and
select the better one. In CLPSO, we de¯ne the ¯tness value the
larger the better, which means that when solving minimization
problems, we will use the negative function value as the ¯tness
values.
3- We use the winner's pbest as the exemplar to learn from for that
dimension. If all exemplars of a particle are its own pbest, we will
randomly choose one dimension to learn from another particle's
pbest's corresponding dimension. The details of choosing fi(d) are
given in (Figure 3).