Meta-heuristic and Evolutionary Algorithms for Engineering Optimization
by Omid Bozorg-Haddad, Mohammad Solgi, Hugo A. Loáiciga
17 Gravity Search Algorithm
Summary
This chapter describes the gravity search algorithm (GSA), an evolutionary optimization algorithm based on the law of gravity and mass interactions. It designates a particle as a solution of an optimization problem. Particles exhibit simple behavior, and they follow intelligent pathways toward the near‐optimal solution. This chapter presents a literature review of the GSA and its applications, explains the GSA’s analogy to the law of gravity and the GSA in detail, and closes with a pseudocode of the GSA.
17.1 Introduction
Rashedi et al. (2009) introduced the gravity search algorithm (GSA) based on the law of gravity and mass interactions and compared it with the particle swarm optimization (PSO) and central force optimization (CFO) with well‐known benchmark functions. Their results established the excellent performance of the GSA in solving various nonlinear functions. Ghalambaz et al. (2011) presented a hybrid neural network and gravitational search algorithm (HNGSA) method to solve the well‐known Wessinger’s equation. Their results showed that HNGSA produced a closer approximation to the analytic solution than other numerical methods and that it could easily be extended to solve a wide range of problems. Jadidi et al. (2013) proposed a flow‐based anomaly detection system and used a multilayer perceptron (MLP) neural network with one hidden layer for solving it. The latter authors optimized the interconnection weights of an MLP network ...