Levy Flight and Chaos Theory-Based Gravitational Search Algorithm for Global Optimization: LCGSA for Global Optimization

Sajad Ahmad Rather, P. Shanthi Bala

Source Title: International Journal of Applied Metaheuristic Computing (IJAMC)13(1)

ISSN: 1947-8283|EISSN: 1947-8291|EISBN13: 9781799885405|DOI: 10.4018/IJAMC.292496

MLA

Rather, Sajad Ahmad, and P. Shanthi Bala. "Levy Flight and Chaos Theory-Based Gravitational Search Algorithm for Global Optimization: LCGSA for Global Optimization." IJAMC vol.13, no.1 2022: pp.1-58. http://doi.org/10.4018/IJAMC.292496

APA

Rather, S. A. & Bala, P. S. (2022). Levy Flight and Chaos Theory-Based Gravitational Search Algorithm for Global Optimization: LCGSA for Global Optimization. International Journal of Applied Metaheuristic Computing (IJAMC), 13(1), 1-58. http://doi.org/10.4018/IJAMC.292496

Chicago

Rather, Sajad Ahmad, and P. Shanthi Bala. "Levy Flight and Chaos Theory-Based Gravitational Search Algorithm for Global Optimization: LCGSA for Global Optimization," International Journal of Applied Metaheuristic Computing (IJAMC) 13, no.1: 1-58. http://doi.org/10.4018/IJAMC.292496

Export Reference

Favorite Full-Issue Download

View Full Text HTML

View Full Text PDF

Abstract

The Gravitational Search Algorithm (GSA) is one of the highly regarded population-based algorithms. It has been reported that GSA has a powerful global exploration capability but suffers from the limitations of getting stuck in local optima and slow convergence speed. In order to resolve the aforementioned issues, a modified version of GSA has been proposed based on levy flight distribution and chaotic maps (LCGSA). In LCGSA, the diversification is performed by utilizing the high step size value of levy flight distribution while exploitation is carried out by chaotic maps. The LCGSA is tested on well-known 23 classical benchmark functions. Moreover, it is also applied to three constrained engineering design problems. Furthermore, the analysis of results is performed through various performance metrics like statistical measures, convergence rate, and so on. Also, a signed Wilcoxon rank-sum test has also been conducted. The simulation results indicate that LCGSA provides better results as compared to standard GSA and most of the competing algorithms.