Comprehensive learning gravitational search algorithm for global optimization of multimodal functions
- PDF / 1,119,033 Bytes
- 36 Pages / 595.276 x 790.866 pts Page_size
- 47 Downloads / 210 Views
(0123456789().,-volV)(0123456789(). ,- volV)
ORIGINAL ARTICLE
Comprehensive learning gravitational search algorithm for global optimization of multimodal functions Indu Bala1 • Anupam Yadav2 Received: 13 July 2018 / Accepted: 9 May 2019 Ó Springer-Verlag London Ltd., part of Springer Nature 2019
Abstract In this paper, a new comprehensive learning gravitational search algorithm (CLGSA) is proposed to enhance the performance of basic GSA. The proposed algorithm is a new kind of intelligent optimization algorithm which has better ability to choose good elements. An intensive comprehensive learning methodology is proposed to enrich the optimization ability of the GSA. The efficiency of the proposed algorithm was evaluated by 28 benchmark functions which have been proposed in IEEE-CEC 2013 sessions. The results are compared with eight state-of-the-art algorithms IPOP, BIPOP, NIPOP, NBIPOP, DE/rand, SPSRDEMMS, SPSO-2011 and GSA. A variety of ways are considered to examine the ability of the proposed technique in terms of convergence ability, success rate and statistical behavior of algorithm over dimensions 10, 30 and 50. Apart from experimental studies, theoretical stability of the proposed CLGSA is also proved. It was concluded that the proposed algorithm performed efficiently with good results. Keywords Gravitational search algorithm Optimization Soft computing Artificial intelligence
1 Introduction Optimization has been a dynamic area of research for many decades. As various real-world optimization problems are becoming complex day by day, better optimization techniques are always required to solve such problems with less cost and less effort. This need forces the topic to be active and challenging to the researchers. Looking back, the research area of optimization algorithms is divided into two major parts—deterministic and non-deterministic or probabilistic, where deterministic approach follows traditional paper–pencil method and nondeterministic approach uses new stochastic techniques to solve optimization problems quickly. Apart from quick solution, the traditional optimization technique fails to & Anupam Yadav [email protected] Indu Bala [email protected] 1
Department of Science and Humanities, Northcap University Gurgaon, Gurgaon, Haryana 122017, India
2
Department of Mathematics, Dr. BR Ambedkar National Institute of Technology Jalandhar, Jalandhar 144011, India
solve NP-hard problems. Therefore, probabilistic approach has been widely used in recent research problems which use optimization. Similarly, central force optimization (CFO) [2, 3] is a deterministic search meta-heuristic for constrained multi-dimensional optimization and gravitational search algorithm (GSA) [1] is a stochastic search algorithm. Both are inspired by particle motion in gravitational field, but due to its stochastic behavior, GSA [1] is more popular than CFO. GSA has been developed in many ways, and various techniques have been used to further improve the search performance of GSA. For instance, Sarafrazi et al. [28] h
Data Loading...
