Genetic Algorithm and its Variants: Theory and Applications

Abstract

The Genetic Algorithm is a popular optimization technique which is bio-inspired and is based on the concepts of natural genetics and natural selection theories proposed by Charles Darwin. The Algorithm functions on three basic genetic operators of selection, crossover and mutation. Based on the types of these operators GA has many variants like Real coded GA, Binary coded GA, Sawtooth GA, Micro GA, Improved GA, Differential Evolution GA. This paper discusses a few of the forms of GA and applies the techniques to the problem of Function optimization and System Identification. The paper makes a comparative analysis of the advantages and disadvantages of the different types of GA. The computer simulations illustrate the results. It also makes a comparison between the GA technique and Incremental LMS algorithm for System Identification.