Improved Krill Herd Algorithm with neighborhood distance concept for numerical optimization

Abstract

Krill herd (KH) is a new recent nature inspired stochastic search algorithm. The fitness function of this algorithm depends upon the minimum distance of each krill from food and the algorithms aims to increase the population diversity around the food cluster. In this paper, the improved krill herd algorithm is employed to solve the optimization problems. The neighborhood distance concept is introduced along with genetic reproduction mechanism in basic KH Algorithm. The results of the three Krill herd algorithms are compared using twelve high dimensional benchmark functions. Along with balancing of exploration and exploitation, computational time of the algorithm is reduced.