Gaussian Bare-Bones Brain Storm Optimization Algorithm

Brain Storm Optimization (BSO) is a population-based algorithm developed based on the humans brainstorming process. It has been successfully applied to many applications in the domain of non-linear continuous optimization. The performance of BSO has been enhanced in the literature through many works attempting to improve its different stages. In this work, we propose a Gaussian Bare-Bones version of the Global-best BSO algorithm (BBGBSO). The idea of bare-bones implementations in general is inspired from the convergence characteristics of Particle Swarm Optimization (PSO) where particles converge to the weighted average of the personal-best of the particle and the global-best of the swarm. A number of previous Bare-bones implementations have been proposed in the literature for different algorithms resulting in noticeable performance improvements. Experimental results extracted from many benchmark functions across different problem sizes confirms the promising performance of BBGBSO.

I. Introduction

Population-based algorithms are a class of algorithms that maintain a population of solutions updated over a number of generations. They have been successfully applied to many nonlinear continuous and/or discrete optimization problems. The population is evolved through a number of iterations until some stopping criterion is met. Population-based algorithms are categorized according to the inspiration behind their evolving strategy. One class is evolutionary algorithms, in which the population update mechanism is developed to mimic the biological evolution process. These algorithms include Genetic Algorithms (GAs), Genetic Programming (GP), Evolutionary Strategies (ES), and Evolutionary Programming (EP). The other class includes swarm intelligence algorithms, in which the population update mechanism is developed to mimic the behavior of living organisms. Some of the algorithms falling under this class are referred to as foraging algorithms as they simulate the foraging behavior of animals and/or insects. Examples of foraging algorithms include Particle Swarm Opti-mization (PSO) [1], [2], Ant Colony Optimization (ACO) [3], Artificial Bee Colony (ABC) [4], and many more. Other swarm intelligence algorithms mimic different kinds of behaviors including for example the egg laying behavior of cuckoos in Cuckoo Search (CS) [5] and the echolocation behavior of bats in the Bat Algorithm (BA) [6].