APPLICATION: FLOOD FORECASTING MODEL FOR KLODZKO CATCHMENT (POLAND) - Modelling Uncertainty in Flood Forecasting Systems - page 95

of genetic information. This is achieved by occasional (with small probability) random
alternation of the value of a string position. In binary coding, this simply means changing
a '1' to a '0' and vice versa.
The offspring obtained from selection—crossover—mutation operations replaces the un-
fit individuals (the solution sets with less fitness values) in the initial population and
forms a new population for next the generation. A general structure of the genetic
algorithms is as follows (Gen and Cheng, 2000):
Procedure: Genetic Algorithms
begin
t ←0;
initialize P (t) ;
evaluate P (t) ;
while (not termination condition) do
begin
recombine P (t) to yield C (t) ;
evaluate C (t) ;
select P ( t +1) from P (t) and C (t) ;
t ← t +1;
end
end
where P(t) is the population for generation t, and C(t) is the offspring produced in
generation t .
5.3.2 GA versions used
Various versions of GA are used in practice. Comprehensive coverage of the topic can be
found in Babovic and Keijzer (2000) and Goldberg (1989) among others. In the present
experiment, two versions of GAs are used:
1. Conventional or normal GA (nGA).
2. Micro-GA (mGA).
The conventional GA is used with tournament selection, uniform crossover, two children
per pair of parents, jump and creep mutations, elitism and niching. The tournament
selection involves both random and deterministic selection features simultaneously. This
method randomly chooses a set of chromosomes and picks up the best chromosomes for
reproduction (Gen and Cheng, 2000). In uniform crossover, given two parent
chromosomes of length 1, each parent copies 1/2 genes to each child, with the selection
of the genes being chosen independently and randomly (Pawlowsky, 1995).