Biomedical Engineering Reference
In-Depth Information
Chapter XIII
Genetic Algorithms and
Multimodal Search
Marcos Gestal
University of A Coruña, Spain
José Manuel Vázquez Naya
University of A Coruña, Spain
Norberto Ezquerra
Georgia Institute of Technology, USA
AbSTRACT
Traditionally, the Evolutionary Computation (EC) techniques, and more specifically the Genetic Algo-
rithms (GAs), have proved to be efficient when solving various problems; however, as a possible lack,
the GAs tend to provide a unique solution for the problem on which they are applied. Some non global
solutions discarded during the search of the best one could be acceptable under certain circumstances.
Most of the problems at the real world involve a search space with one or more global solutions and
multiple local solutions; this means that they are multimodal problems and therefore, if it is desired to
obtain multiple solutions by using GAs, it would be necessary to modify their classic functioning outline
for adapting them correctly to the multimodality of such problems. The present chapter tries to establish,
firstly, the characterisation of the multimodal problems will be attempted. A global view of some of the
several approaches proposed for adapting the classic functioning of the GAs to the search of multiple
solutions will be also offered. Lastly, the contributions of the authors and a brief description of several
practical cases of their performance at the real world will be also showed.
INTRODUCTION
a population: a randomly generated initial set of
individuals. Every one of these individuals -who
represent a potential solution to the problem-,
will evolve according to the theories proposed by
Darwin (Darwin, 1859) about natural selection
Following a general prospect, the GAs (Holland,
1975) (Goldberg, 1989) try to find a solution using
Search WWH ::
Custom Search