Author(s)

B. Dilimulati & I. Bruha

Abstract

Genetic Algorithms (GAs) are search methods based on principles of natural selection and genetics. GAs attempt to find optimal solutions to a given problem by manipulating a population of candidate solutions (individuals). In the real world, we always encounter the problems that need to be solved in a changing environment. This means that our algorithm needs to be dynamic or even adaptive to the changing environment. In this paper, we mainly deal with the adaptive GAs that have a new genetic operator called transformation instead of the traditional crossover. We use a dynamic problem generator to create a dynamically changing landscape and study the behavior of the transformation-based GAs in different parameter settings, such as transformation rate, mutation rate and segment replacement rate. Keywords: Genetic Algorithm, transformation operator, dynamically changing environment. 1 Introduction Genetic Algorithms (GAs) are mainly used to solve optimization problems, see e.g. [1, 3, 4, 7]. In fact, there are various optimization methods such as exhaustive search, analytical optimization, line optimization methods, and natural optimization methods; natural optimization methods include simulated annealing, ant colony optimization, and genetic algorithms. In traditional GAs, the operator set is usually fixed but in the real world we always encounter problems that need to be solved in a changing environment. Such problems include target recognition (the sensor performance varies on environmental conditions); scheduling problems (available resources vary over

Keywords

Genetic Algorithm, transformation operator, dynamically changing environment.