Gravitational Search Algorithm

Abstract

Today, many metaheuristics algorithms have been developed are inspired by the physical phenomena or behaviors of natural creatures that are very effective in solving complex engineering optimization problems. These algorithms seek to find optimal solutions in various engineering problems such as machine learning, image processing, pattern matching, decision making, and so on. These algorithms can be implemented in both single-objective and multi-objective optimization problems. Among the most famous techniques in this category include the ant colony optimization (ACO) algorithm, which is designed based on the navigation of ants; particle swarm optimization (PSO), which is based on the movement of organisms in a bird flock; and gravitational search algorithm (GSA). GSA is one of the most popular metaheuristics methods. It is inspired by the law of gravity and motion and creating the appropriate balance between exploration and exploitation capabilities. In this chapter, we are going to introduce the basic GSA and review the newer versions. First, we will introduce the original GSA that is proposed for continuous problems. Then we will discuss other versions of GSA for various optimization problems and analyze the convergence properties of this algorithm. We will also review the GSA in various engineering applications.