UNIFORM POPULATION IN GENETIC ALGORITHMS

The most of researchers dealing with genetic algorithms apply variation on the genetic operators. Some of them propose new genetic algotihm types. However, none of them deals with generating population of good quality, initially. In this paper, we propose a method for generating initial population and the method includes all types of chromosome encoding. The goodness of generated population by proposed method is also illustrated by applying this population and random initial population to multi-modal functions such as Griewank, Michalewicz and Rastrigin. For the sake of simplicity, all functions are selected as functions of two variables.